Maurice A. Deane School of Law at Hofstra UniversityScholarly Commons at Hofstra Law

Hofstra Law Faculty Scholarship

1986

Between Two Worlds: The Shift from Individual toGroup Responsibility in the Law of Causation ofInjuryRobert A. Baruch BushMaurice A. Deane School of Law at Hofstra University

Follow this and additional works at: https://scholarlycommons.law.hofstra.edu/faculty_scholarship

This Article is brought to you for free and open access by Scholarly Commons at Hofstra Law. It has been accepted for inclusion in Hofstra Law FacultyScholarship by an authorized administrator of Scholarly Commons at Hofstra Law. For more information, please contact lawcls@hofstra.edu.

Recommended CitationRobert A. Baruch Bush, Between Two Worlds: The Shift from Individual to Group Responsibility in the Law of Causation of Injury, 33UCLA L. Rev. 1473 (1986)Available at: https://scholarlycommons.law.hofstra.edu/faculty_scholarship/674

ARTICLES

BETWEEN TWO WORLDS: THE SHIFTFROM INDIVIDUAL TO GROUP

RESPONSIBILITY IN THE LAW OFCAUSATION OF INJURY

Robert A. Baruch Bush*

The principle of individual responsibility is one of thefoundations of tort law, and indeed of the common law ingeneral. Recently, however, court decisions and scholarlycommentary, particularly on the issue of actual causation ofinjury, indicate a shift away from the principle of individualresponsibility in torts toward what can be termed group, orcommunity, responsibility. While the dimensions of thisshift are as yet limited, its implications are immense: it callsinto question fundamental premises regarding the nature ofthe individual and the individual's relationship to society.This Article addresses the emerging shift in tort law fromindividual to group responsibility, and its legal, political, andphilosophical implications. After tracing the shift anddescribing the generally confused and hostile responses that

* Associate Professor of Law, Hofstra Law School. B.A., Harvard Univer-

sity, 1969; J.D., Stanford Law School, 1974.

The research for this Article was supported in part by a grant from the Hof-stra University School of Law, for which I am grateful. I also thank the followingcolleagues for their many helpful comments and suggestions on this Article:Linda Champlin, David Diamond, Bernard Jacob, John Gregory, Janet Dolgin,Lawrence Kessler, and, especially, Aaron Twerski and Lawrence Joseph. MarcRoss, Christopher Flint, Howard Poliner, Dina Epstein, and John Bernstein pro-vided valuable research assistance. Special thanks for helping to uncover invalua-ble source materials are due to Mrs. David Bush, the author's mother. Finally, thegreatest debt and the deepest thanks are owed to the author's constant partner indiscussing and developing the ideas expressed herein-his wife, Dr. Susan E. Shu-lamis Bush.

1473

UCLA LA W REVIEW

it has evoked, I demonstrate that the shift is justified by, andintuitively expresses, a revision of the political theory of theindividual and society that underlies present conceptions oflegal responsibility. I conclude that the shift to group re-sponsibility is a positive development, but one that must beunderstood and applied much more critically than courts orcommentators have done thus far.

INTRODUCTION:

THE INDIVIDUAL RESPONSIBILITY PRINCIPLE IN

TORT LAW AND ITS CHALLENGER

According to the individual responsibility (IR) principle,an individual is responsible for the consequences of his ownactions: that is, (1) responsibility for those consequencesrests on him and not on some other(s) who had no part inbringing them about; (2) he bears responsibility only for hisactions and not for the actions of some other(s) in which heplayed no part. In other words, the individual is responsiblefor all he does, but for only what he does. This is the essenceof the IR principle as it has operated in tort law.

This principle underlies both of the major elements of atypical torts case-negligence and actual causation. Negli-gence is assessed on the basis of whether the individual de-fendant's conduct, and not someone else's, met or deviatedfrom the required standard of conduct.' Likewise, actualcausation requires a showing that the individual defendant'sconduct, and not someone else's, was a necessary link in thechain of events that caused the plaintiff's harm.2 This is at

1. See W. PROSSER & P. KEETON, PROSSER AND KEETON ON THE LAW OF TORTS§ 30, at 164-69, § 39, at 248-49 (5th ed. 1984); see also Newing v. Cheatham, 15Cal. 3d 351, 363, 540 P.2d 33, 41, 124 Cal. Rptr. 193, 201 (1975) (court explainedthat the purpose of the exclusive control component of the res ipsa loquitur doc-trine was "to link the [individual] defendant to the probability . . . that the acci-dent was negligently caused"); Sheffield v. Eli Lilly & Co., 144 Cal. App. 3d 583,596-98, 192 Cal. Rptr. 870, 877-78 (1983) (court rejected plaintiff's argumentthat several vaccine manufacturers should be held liable, since plaintiff's injurieswere caused by a single defectively produced batch of vaccine and as a result onlyone of the manufacturers was guilty of "tortious conduct," i.e., negligence).

2. See R. HURSH & H. BAILEY, AMERICAN LAW OF PRODUCTS LIABILITY § 1.41(2d ed. 1974); W. PROSSER & P. KEETON, supra note 1, § 41, at 263-65, 269; Zwier,"Cause in Fact" in Tort Law: A Philosophical and Historical Examination, 31 DE PAUL L.REV. 769, 777-79 (1982); see also Ryan v. Eli Lilly & Co., 514 F. Supp. 1004, 1017(D.S.C. 1981); Payton v. Abbott Laboratories, 386 Mass. 540, 571-73, 437 N.E.2d171, 188-89 (1982); Namm v. Charles E. Frosst & Co., 178 N.J. Super. 19, 28, 427

1474 [Vol. 33:1473

1986] BETWEEN TWO WORLDS 1475

the basic doctrinal level.At a broader level, the IR principle seems to lie at the

foundation of the two theories most commonly offered toexplain the social function of tort law: the utilitarian theoryand the corrective justice theory.3 In the utilitarian view,

A.2d 1121, 1125 (App. Div. 1981). There is no contradiction between this re-quirement and the doctrine ofjoint and several liability. See infra text accompany-ing notes 6-9. In a strict liability regime, where negligence need not be shown,actual causation is the central requirement for liability; therefore, the IR principleencompasses both negligence and strict liability. Thus, the controversial shift intort law from negligence to strict liability, if it has occurred, has not displaced theIR principle.

Some contemporary writers argue that the actual causation requirement doesnot represent any kind of objective measure of the individual defendant's involve-ment in an injury. They suggest instead that all questions of causation are ulti-mately policy questions and are answerable only on nonobjective, politicalgrounds. See, e.g., Horowitz, The Doctrine of Objective Causation, in THE POLITICS OFLAw: A PROGRESSIVE CRITIQUE 201 (D. Kairys ed. 1982). However, Horowitz'sargument and historical discussion deal essentially with the proximate cause ques-tion and its concern for choosing among multiple targets for liability. It is truethat no one today accepts the notion of identifying a sole proximate cause of in-jury by objective means free of policy considerations. However, the actual causequestion is a different matter. Before courts will even begin the policy-orientedproximate cause inquiry, they require a showing that the defendant is at least anactual cause of the plaintiff's injury. And courts usually treat this actual causationquestion as a policy-neutral, factual inquiry. See Wright, Causation in Tort Law, 73CALIF. L. REV. 1735, 1740-41 (1985). Wright points out that the "policy-cause"theorists generally confuse the actual causation issue with the policy-oriented neg-ligence or proximate cause issues, and shows at length how "the causal inquiry isa factual, empirical issue that can be-and almost always is-kept distinct from thepolicy issues in tort adjudication." Id. at 1803; see also id. at 1803-13. On theother hand, this Article argues that the actual causation requirement itself is not aneutral concept, but a manifestation of a particular world view associated with aparticular theory of society. See infra text accompanying notes 156-86.

3. See G. WHITE, TORT LAw IN AMERICA 218-30 (1980); Wright, Actual Causa-tion vs. Probabilistic Linkage: The Bane of Economic Analysis, 14 J. LEGAL STUD. 435,435-37 (1985). The utilitarian theory takes the view that tort law's function is tocreate incentives for individuals to use resources efficiently, that is, in a fashionthat maximizes societal welfare. See, e.g., Fletcher, Fairness and Utility in Tort Law,85 HARV. L. REV. 537, 537-38, 566-69 (1972); Hutchinson, Beyond No-Fault, 73CALIF. L. REV. 755, 758-59 (1985); Rosenberg, The Causal Connection in Mass Expo-sure Cases: A "Public Law" Vision of the Tort System, 97 HARV. L. REV. 851, 859,861-62 (1984). The corrective justice theory takes the view that tort law's func-tion is to correct individual injustices. See, e.g., Englard, The System Builders: A Criti-cal Appraisal of Modern American Tort Theory, 9 J. LEGAL STUD. 27, 27 (1980);Rosenberg, supra, at 859-60, 877; Wright, supra, at 435; see also infra notes 4-5 andaccompanying text.

I adopt the terms "utilitarian" and "corrective justice" because they arewidely used and understood. However, different labels are often used for eachtheory. For example, the utilitarian theory is also called the "functional" theory,see Englard, supra, at 32-34, the "economic" theory, see Posner, A Theory of Negli-gence, 1J. LEGAL STUD. 29, 32-33 (1972); Wright, supra, at 436-39, and the "deter-

UCLA LA W REVIEW

tort law exists to create incentives for social-cost-minimizingbehavior.4 Since individuals are the targets of these incen-tives, liability must be based on individual conduct, or it willdistort incentives and lead to socially suboptimal behavior.In the corrective justice view, tort law exists to correct injus-tices committed by one individual against another; there-fore, liability clearly must be based on individual conduct. 5

Thus, the IR principle is fundamental to tort law at both thedoctrinal and theoretical levels.

Of course, there are situations where more than one in-dividual is held responsible to an injured victim. When theactions of several individuals combine, either sequentially orcontemporaneously, to produce a particular harm thatwould not occur in the absence of any one of them, all areheld jointly and severally liable as tortfeasors.6 However,under joint and several liability, each tortfeasor is still seenas individually responsible to the victim for the entire harm.The separate rights of contribution or indemnity that eachtortfeasor might have against the others7 do not change thefact that each one may have to pay the entire judgment. Nordoes this result violate the IR principle. Since eachtortfeasor's actions constitute a cause sine qua non of the en-tire harm, each can properly be held fully responsible. Inother cases, a number of individuals are held liable for hav-ing "acted in concert" in a way that resulted in the victim'sinjury.8 Again, the IR principle remains intact. Each indi-

rence" or "deterrent" theory of tort law, see G. CALABRESI, THE COSTS OFACCIDENTS 26-27, 68-69 (1970); Calabresi, Optimal Deterrence and Accidents, 84YALE L.J. 656, 656-57 (1975); Rosenberg, supra, at 861-62; Note, Market ShareLiability for Defective Products: An Ill-Advised Remedy for the Problem of Identification, 76Nw. U.L. REV. 300, 315-21 (1981).

4. See, e.g., Posner, supra note 3, at 32-33; Note, supra note 3, at 315-21.5. See Englard, supra note 3, at 27; Wright, supra note 3, at 435. Beyond the

functional importance of individual responsibility to both theories, the two arealso deeply committed at their philosophical roots to the concepts of the integrity,autonomy and inviolability of the individual, which any principle but individualresponsibility would negate. See infra notes 170, 186.

6. See W. PROSSER & P. KEETON, supra note 1, § 41, at 268, 271-72, § 47, at326, § 52, at 345-48; see also Michie v. Great Lakes Steel Div., Nat'l Steel Corp.,495 F.2d 213 (6th Cir.), cert. denied, 419 U.S. 997 (1974); Hackworth v. Davis, 87Idaho 98, 390 P.2d 422 (1964); Maddux v. Donaldson, 362 Mich. 425, 108N.W.2d 33 (1961); Zulauf v. New York, 119 Misc. 2d 135, 462 N.Y.S.2d 560 (Ct.Cl. 1983), affd, 110 A.D.2d 1042, 489 N.Y.S.2d 1019 (1985); Landers v. East Tex.Salt Water Disposal Co., 151 Tex. 251, 248 S.W.2d 731 (1952).

7. See W. PROSSER & P. KEETON, supra note 1, §§ 50-51, at 336-45.8. See id. § 46, at 322-24; RESTATEMENT (SECOND) OF TORTS § 876 (1965); see

1476 [Vol. 33:1473

BETWEEN TWO WORLDS

vidual is considered a necessary link in a "conspiratorial"circle, and each therefore is a cause sine qua non and bearsindividual responsibility to the victim for the entire harm.

The longstanding doctrine of vicarious liability is moredifficult to reconcile with the IR principle. 9 For example, ifan employer can be held fully liable for harm caused by anemployee without any proof of dereliction in the supervisionof the employee, the result would seem to violate the IRprinciple. In reality, it may; however, courts have labored toshow otherwise. Frequently invoking the presumed agencyrelationship between the employer and employee, courtshave insisted that the employer has acted individuallythrough the employee to injure the victim.' 0 The tenuous-ness of this argument attests to the commitment of its au-thors to regard the IR principle as sacrosanct.

Nevertheless, the doctrine of vicarious liability doessuggest an alternative to the IR principle. Professor Fein-berg, who has written extensively on the philosophy of legalresponsibility, discusses the possibility of imposing vicari-ous, or collective, liability on a group for the actions of itsconstituent members: "When the whole group .. .is heldresponsible for the actions of one or some of its members,then, from the point of view of any given responsible indi-vidual, his liability will in most cases be vicarious."'" Whatemerges from Feinberg's discussion is the principle of groupresponsibility. This principle would hold an individual re-sponsible for others' actions, and vice-versa, not becausethey actually acted together, but simply because they are allassociated within a common group. Thus, a victim could ar-gue, "since I can show that I was injured by someone in thisgroup, everyone in this group is obligated to pay me damages,whether or not he was, or acted with, the actual injurer."Such an argument is plainly incompatible with the IR princi-ple. Therefore, to accept this argument openly would be tochallenge the IR principle and the political-philosophical ba-sis on which it rests. Even the bare form of this argument is

also Orser v. George, 252 Cal. App. 2d 660, 60 Cal. Rptr. 708 (1967); Agovino v.Kunze, 181 Cal. App. 2d 591, 5 Cal. Rptr. 534 (1960).

9. See W. PROSSER & P. KEETON, supra note I, § 69, at 499-501.10. See id. at 499-500; see, e.g., Ives v. South Buffalo R.R., 201 N.Y. 271, 94

N.E. 431 (1911).11. J. FEINBERG, DOING AND DESERVING 233 (1970) (emphasis in original).

1986] 1477

UCLA LA W REVIEW

likely to offend some of our deeply held ideas about respon-sibility by suggesting "guilt by association" and "collectiveguilt."

Perhaps for these reasons, few have openly advocatedadoption of the group responsibility principle. However,courts and commentators in a growing number of cases andarticles have actually introduced what can only be under-stood as group responsibility.' 2 In doing so, these courtsand scholars have neither advocated nor even recognizedthe principle of group responsibility: they have justifiedtheir decisions and proposals by reference to the rationalesof utilitarianism and corrective justice, both of which may belinked to the IR principle itself.'3 As a result, these deci-sions and commentaries are badly confused and internallyinconsistent. Nevertheless, many critics have recognized theimplication of these cases and proposals and have stronglyresisted what they rightly see as a challenge to the IRprinciple. 14

The time has come to begin openly the debate betweenthe principle of individual responsibility and that of groupresponsibility, and the political-philosophical premises thatunderlie each. This Article attempts to open the debate inthe area of tort law, just as it has been opened in otherfields. 15 Only by doing so can we make intelligent choices

12. See, e.g., Sindell v. Abbott Laboratories, 26 Cal. 3d 588, 607 P.2d 924, 163Cal. Rptr. 132, cert. denied, 449 U.S. 912 (1980); Abel v. Eli Lilly & Co., 418 Mich.311, 343 N.W.2d 164, cert. denied, 469 U.S. 833 (1984); Bichler v. Eli Lilly & Co.,79 A.D.2d 317, 436 N.Y.S.2d 625 (1981), aff'd, 55 N.Y.2d 571, 436 N.E.2d 182,450 N.Y.S.2d 776 (1982); Collins v. Eli Lilly & Co., 116 Wis. 2d 166, 342 N.W.2d37, 469 U.S. 826 (1984); Delgado, Beyond Sindell: Relaxation of Cause-In-Fact Rules

for Indeterminate Plaintiffs, 70 CALIF. L. REV. 881 (1982); King, Causation, V1aluation,and Chance in Personal Injury Torts Involving Preexisting Conditions and Future Conse-quences, 90 YALE L.J. 1353 (1981); Robinson, Multiple Causation in Tort Law: Reflec-tions on the DES Cases, 68 VA. L. REV. 713 (1982); Rosenberg, supra note 3.

13. See supra notes 3-5 and accompanying text.14. See infra notes 111-34 and accompanying text.15. For examples of this debate in other fields, including philosophy,

medicine, and political science, seeJ. FEINBERG, supra note 11; M. SANDEL, LIBER-ALISM AND ThE LIMITS OF JUSTICE (1982); French, Collective Responsibility and thePractice of Medicine, 7J. MED. & PHIL. 65 (1982); Weale, Representation, Individualism,and Collectivism, 91 ETHICS 457 (1981).

In tort law itself, some commentators, particularly from the Critical LegalStudies school, have more or less openly challenged the IR principle as a basis forimposing liability or awarding recovery. See, e.g., Abel, Torts, in THE POLITICS OFLAW, supra note 2, at 185; Abel, A Socialist Approach to Risk, 41 MD. L. REV. 695(1982) [hereinafter cited as Abel, Socialist Approach]; Horowitz, supra note 2;

1478 [Vol. 33:1473

1986] BETWEEN TWO WORLDS 1479

about tort law, instead of wandering between two worlds,with some stumbling along in confusion and others settlinginto blind resistance to change.

Part I of this Article documents the emerging shift fromindividual to group responsibility in tort law, focusing on theactual causation issue. Part II describes the prevailing atti-tudes toward the shift, which range from resignation toresistance, and then demonstrates that none of these re-sponses views group responsibility as a positive phenome-non. This hostility to group responsibility stems not froman attachment to the IR principle per se, but from a commit-ment to the underlying political theory of liberalism, whichdominates modern American political thought, and from acorresponding aversion to the rival political theory of socialwelfarism, which appears to underlie group responsibility.1 6

Part III suggests that in introducing group responsibility,

Hutchinson, supra note 3. These critics object to the IR principle and the tradi-tional causation requirement on the ground, among others, that these conceptsare tools of "liberal legalism" that are used "to protect ruling elites from theclaims of those they oppress." Levinson, Book Review, 96 HARV. L. REV. 1466,1482 (1983) (reviewing THE POLITICS OF LAw, supra note 2). However, their pro-posed alternatives rarely go beyond socialized insurance, see Abel, Torts, supra, at196-200, and advocacy of "participatory democracy" in local health and safetyactivities, see Hutchinson, supra note 3, at 770-71. Neither of these alternativeswould involve group responsibility as discussed in this Article. Hence, these crit-ics have not yet specifically or directly addressed the debate I am opening here.As Levinson puts it, in commenting on Abel's work, he "does not begin to de-scribe a nonindividualist conception of injury that would allow us to transcendliberal legalism." Levinson, supra, at 1482-83.

16. Liberalism is also known as liberal individualism or the liberal theory; so-cial welfarism is also known as welfarism or the social welfare theory. The liberaltheory takes the view that the individual is the fundamental unit and basis of soci-ety and society's sole function and value is to facilitate self-fulfillment for individu-als. See M. SANDEL, supra note 15, at 1-11, 50-53, 66-67; Murphy, Liberalism andPolitical Society, 26 AM.J.JuRIS. 125, 125, 153 (1981). This political theory is asso-ciated with the corrective justice theory of torts. See infra text accompanying notes157-86. The social welfare theory takes the view that the good of society in theaggregate is the supreme value and that this good can and should override indi-vidual rights and fulfillment when the two values conflict. See Rosenberg, supranote 3, at 859-61; Wright, supra note 3, at 436. This political theory is associatedwith the utilitarian theory of torts. See infra text accompanying notes 157-86.

There is less consensus on terminology in political theory than in tort theory.See supra note 3. Liberalism is also referred to as "rights-based" theory, see Hutch-inson, supra note 3, at 759-60, "individual rights" or "individualist" theory, seeEnglard, supra note 3, at 32-33, 57-68; Rosenberg, supra note 3, at 859-60, and"fairness" theory, see Fletcher, supra note 3, at 537-43. Some scholars subdivideliberalism into "rights-based" and "libertarian" theory. See Hutchinson, supranote 3, at 757-60. Social welfarism is also referred to as political "utilitarianism"and "collectivism." See M. SANDEL, supra note 15, at 3-5, 50-51, 66-67, 138-47,

1480 UCLA LAW REVIEW [Vol. 33:1473

courts have actually not embraced social welfarism but haverather intuited the outlines of an alternative vision to bothliberalism and social welfarism. This alternative vision, re-ferred to as the communitarian theory of society,' 7 possessesa number of advantages over both liberal and welfarist the-ory. In addition, it both justifies and rationally restricts theapplication of group responsibility in torts. My conclusion isthat courts should openly encourage the shift to group re-sponsibility, but only if they explicitly base this responsibilityon communitarian grounds. For only the communitariantheory can justify and properly restrict this alternative prin-ciple of tort law.

I. THE SHIFT FROM INDIVIDUAL TO GROUP RESPONSIBILITY

IN THE LAW OF CAUSATION OF INJURY

A. The Actual Causation Requirement

Traditionally, a basic requirement for recovery in anytort action, whether based on negligence or strict liability,has been that the plaintiff must prove by a preponderance ofthe evidence that the defendant's conduct was an actual

165-67; Rosenberg, supra note 3, at 859, 907; Sandel, Morality and the Liberal Ideal,NEW REPUBLIC, May 7, 1984, at 15, 17.

My choice of "liberalism" and "social welfarism" as descriptive terms rests onseveral grounds. In political theory itself, the term "liberalism" is the acceptedusage to describe the individualist, rights-based vision. The term "social welfar-ism," while it is used less widely, is less confusing in the context of the presentArticle than "utilitarianism" or "collectivism." The term "utilitarianism," in thepolitical theory context, has its own historical background in nineteenth-centuryindividualist political philosophy. Consequently, the term might confuse thereader were it used to describe a theory based on aggregate welfare. See Brinton,Utilitarianism, in 15 ENCYCLOPEDIA OF THE SOCIAL SCIENCES 197 (E. Seligman ed.1935). Furthermore, the use of a term different than that used to describe theassociated tort theory is helpful to distinguish whether tort law or political philos-ophy is under discussion. The term "collectivism" is, in my view, simply too ideo-logically loaded to be useful as a primary descriptive expression. Given theforegoing clarification, it should be clear that "liberalism" does not refer to polit-ical policies "on the left" and that "social welfarism" does not refer to govern-mental programs designed to aid the disadvantaged. While some possibility ofconfusion may remain, the attentive reader will easily avoid it, and the choice ofthe liberalism/social welfarism terminology both reflects the literature of politicaltheory and avoids more serious confusion.

17. The term "communitarian" is taken from current work in political philos-ophy. See M. SANDEL, supra note 15, at 147-54, 172-74; Sandel, supra note 16, at17. Its meaning is developed in depth in the text infa accompanying notes188-234.

BETWEEN TWO WORLDS

cause of his injury.' 8 This legal rule is directly related to theIR principle.' 9 It requires a showing that the plaintiff'sharm was a consequence of the defendant's actions, so thatliability makes the defendant responsible only for his ownactions. 20 It demands that the victim identify the party whowas individually responsible for his injury.21 Despite thiscongruence between the traditional causation requirementand the IR principle, it is precisely in the causation area thatthe shift from individual to group responsibility has beenmost pronounced.

In the causation context, group responsibility takes twoforms: holding individuals responsible for each other's inju-rious actions when all are members of a common group, andallowing individuals to recover for each other's injurieswhen all are members of a common group. This Part showshow a number of court decisions and scholarly proposalshave adopted group responsibility in one or both of theseforms, although few if any directly acknowledge it.

B. Case Law: "Special" Causation Rules for Multiple iCausal Possibility

In a number of recent decisions, courts have allowedplaintiffs to proceed even when they could not possibly sat-isfy the traditional causation requirement.22 Several of these

18. See Namm v. Charles E. Frosst & Co., 178 N.J. Super. 19, 28, 427 A.2d1121, 1125 (App. Div. 1981); W. PROSSER & P. KEETON, supra note 1, § 41, at 269;Zwier, supra note 2, at 777-79.

19. See supra notes 1-2 and accompanying text.20. See Payton v. Abbott Laboratories, 386 Mass. 540, 571-73, 437 N.E.2d

171, 188-89 (1982).21. See id. at 571, 437 N.E.2d at 188; Zwier, supra note 2, at 777-79. Of

course, more than one such party may be identified in cases of joint tortfeasors.Functionally, the actual causation requirement furthers the utilitarian goals of tortlaw. It ensures that the responsible party cannot escape the full burden of liabilityby shifting blame to another. It also avoids placing any burden on actors, evenblameworthy ones, whose conduct was in fact harmless. The full force of deter-rence is thus directed at individuals who actually produce harm. Individuals whodo not actually produce harm are spared from "overdeterrence," so that incen-tives for cost-minimizing behavior are clear and undistorted. See Posner, supranote 3, at 40-42. At the same time, the identification requirement is crucial to thefulfillment of the corrective justice goal. Only after identifying the injurer is itpossible to correct the specific injustice done, by having the individual injurer,and no one else, repay his individual victim.

22. See, e.g., Hardy v. Johns-Manville Sales Corp., 509 F. Supp. 1353 (E.D.Tex. 1981), rev'd, 681 F.2d 834 (5th Cir. 1982); Sindell v. Abbott Laboratories, 26Cal. 3d 588, 607 P.2d 924, 163 Cal. Rptr. '132, cert. denied, 449 U.S. 912 (1980);

1986] 1481

UCLA LA W REVIEW

cases presented a similar fact pattern: plaintiffs injured by agenerically marketed drug, DES, could not identify the spe-cific manufacturer of the DES that injured them. The deci-sions in these cases and certain others have varied in theirapproaches and rationales and have articulated a number of"special" cause-in-fact rules to deal with the situationspresented. All these "special" rules have implicitly intro-duced the group responsibility principle.

1. Concert of Action

A court may hold several tortfeasors jointly and sever-ally liable when they have acted in concert in a way thatcauses a plaintiff injury. 23 This does not contradict the IRprinciple. As long as "concerted action" is defined as in-volving mutual cooperation and encouragement by each de-fendant, then each can be considered a necessaryparticipant. Hence, each is individually responsible for theentire harm. However, at least one court in a DES case hasheld that a plaintiff can use a concerted action argument toestablish causation even when the alleged concert of actioninvolved no more than "conscious parallel" behavior.

In Bichler v. Eli Lilly & Co., 24 the court concluded that, intesting and marketing DES, each of the multiple defendantsacted independently, but in a manner that each knew wassimilar to the others' actions. 25 By recognizing a distinctionbetween this kind of behavior and traditional concert of ac-tion, the court indicated that it did not believe that the de-fendants had "acted as one" or had "aided and abetted" oneanother such that each could be considered individually re-sponsible for the entire harm. Nevertheless, while acknowl-edging that each manufacturer had acted independently, thecourt still held each liable to the plaintiff for one hundred

Copeland v. Celotex Corp., 447 So. 2d 908 (Fla. Dist. Ct. App. 1984), rev 'd in partand modified, 471 So. 2d 533 (Fla. 1985); Abel v. Eli Lilly & Co., 418 Mich. 311, 343N.W.2d 164, cert. denied, 469 U.S. 833 (1984); Bichler v. Eli Lilly & Co., 79 A.D.2d317, 436 N.Y.S.2d 625 (1981), aff'd, 55 N.Y.2d 571, 436 N.E.2d 182, 450N.Y.S.2d 776 (1982); Martin v. Abbott Laboratories, 102 Wash. 2d 581, 689 P.2d368 (1984); Collins v. Eli Lilly & Co., 116 Wis. 2d 166, 342 N.W.2d 37, cert. denied,469 U.S. 826 (1984).

23. See supra text accompanying note 8.24. 79 A.D.2d 317, 436 N.Y.S.2d 625 (1981), aff'd, 55 N.Y.2d 571, 436

N.E.2d 182, 450 N.Y.S.2d 776 (1982).25. Id. at 324-26, 329, 436 N.Y.S.2d at 630-31, 632.

[Vol. 33:14731482

BETWEEN TWO WORLDS

percent of the damages. In effect, since the plaintiff was un-able to point a finger at a responsible individual, the courtinstead allowed her to point to a group, which it defined bysimilarity of behavior. The court then held each member ofthe group fully responsible for what may have been, and insome cases certainly was, other members' actions-effec-tively adopting the group responsibility principle.

2. Alternative Liability

Earlier cases that dealt with the problem of impossibilityof proof of causation provided another basis for allowing re-covery in DES cases. The alternative liability approachoriginated in the often cited case of Summers v. Tice. 26 In thiscase, the plaintiff had been injured by a bullet that certainlywas fired by one of two negligent hunters; however, which ofthe two had fired the bullet could not be determined. Re-jecting a concert-of-action rationale, the court decided forreasons of "practical justice" to shift the burden of proof onthe issue of causation and to hold each defendant fully liableif he could not exonerate himself.27 However, shifting theburden of proof actually changed the plaintiff's burden fun-damentally and permanently. Once shifted, the burdencould not be overcome in Summers, nor in most other cases inwhich Summers has been applied. In effect, the Summers ruleis that the plaintiff who is unable to point to a single negli-gent actor as responsible can point instead to a group ofnegligent actors that certainly includes the responsible indi-vidual. All members of that group are liable, each for theentire harm.

The Summers approach, dubbed "double fault and alter-native liability" by Dean Prosser,28 has been applied subse-quently in a variety of group settings, from medicalmalpractice 29 to defective products 30 to multiple-car colli-sions 3 1 and finally to DES cases. 32 The key element of theSummers rationale in all of these situations, including Summers

26. 33 Cal. 2d 80, 199 P.2d 1 (1948).27. See id. at 86-88, 199 P.2d at 4-5.28. W. PROSSER & P. KEETON, supra note 1, § 41, at 270-71; see RESTATEMENT

(SECOND) OF TORTS § 433B(3) (1965).29. See Ybarra v. Spangard, 25 Cal. 2d 486, 154 P.2d 687 (1944).30. See Anderson v. Somberg, 67 N.J. 291, 338 A.2d 1 (1975).31. See Vahey v. Sacia, 126 Cal. App. 3d 171, 178 Cal. Rptr. 559 (1981); Cop-

ley v. Putter, 93 Cal. App. 2d 453, 207 P.2d 876 (1949).

1986] 1483

UCLA LA W REVIEW

itself, is the courts' decision to hold all the individuals in agroup responsible because that group certainly includes the"guilty" party. This means that all but one in the group willpay for some other member's actions-a shift from individ-ual to group responsibility.

3. "Market Share" Liability

"Market Share" liability is a third theory used to allowDES plaintiffs to proceed despite their inability to provewhich defendant caused their injury. This theory, adoptedby the California Supreme Court in Sindell v. Abbott Laborato-ries,33 is best understood as an attempt to balance a concernfor plaintiff's recovery against a perception that the concertof action and alternative liability approaches were unfair todefendants. 34 The court fashioned a compromise. Plaintiffcould point to a group of possible causal actors that proba-bly, but not certainly, included the actual responsible party.Each member of this group would be liable unless exoner-ated, but the liability of each would be limited to a portion ofplaintiff's damages proportional to that defendant's share ofthe total sales of the drug. In effect, the market share repre-sented the rough equivalent of the statistical probability ofcausation by each defendant.3 5 This approach diluted theSummers requirement that all possible causal actors bejoined, making it easier for plaintiff to recover; but at thesame time it limited individual exposure, making it less likelythat individual defendants would bear full liability. As acompromise between no recovery and joint and several lia-bility, market share liability was brilliant. However, it is aneven clearer example than the two previously discussed"special" causation rules of the shift from individual togroup responsibility.

32. See Abel v. Eli Lilly & Co., 418 Mich. 311, 343 N.W.2d 164, cert. denied, 469U.S. 833 (1984).

33. 26 Cal. 3d 588, 607 P.2d 924, 163 Cal. Rptr. 132, cert. denied, 449 U.S. 912(1980).

34. Under either of those approaches, any single defendant included withinthe group responsible for the harm could bear the entire burden of the plaintiff'sdamages. The Sindell court was probably troubled about imposing such extensiveliability for a mere possibility of causation. However, the court was equally trou-bled by the solution of barring the plaintiff's recovery unless causation could beproven in the traditional manner.

35. Sindell, 26 Cal. 3d at 612-14, 607 P.2d at 937-38, 163 Cal. Rptr. at145-46.

1484 [Vol. 33:1473

BETWEEN TWO WORLDS

The Sindell court, implicitly recognizing the problem ofadopting what amounted to group responsibility, empha-

sized that under its theory, each defendant would pay for nomore than the aggregate injuries caused by its own actions. 36

In other words, the court claimed that the IR principle,threatened by the other theories, was kept intact here. How-ever, this was disingenuous. Sindell's market share liabilityrepresents group responsibility in several ways. First, as thedissent in Sindell and other observers have argued, the deci-sion implies that the plaintiff who sues can recover one hun-dred percent of her damages from the named defendants,even if they represent less than one hundred percent of themarket, provided that they represent a substantial share.37 If

this reading is accurate, then even considering aggregateharm, each individual defendant would almost certainly pay

for harm done by others-that is, producers in the samemarket who are not or cannot be joined as defendants in thelitigation. Second, even if all producers were joined in aSindell action, the strong possibility remains that not everyvictim would be included, whether in a class suit or a series ofindividual suits. If this is the case, some of the defendantswho pay on a market share basis once again will pay forharm they did not do to the actual plaintiffs, even in the ag-gregate, since their victims, by hypothesis, would not beplaintiffs at all. Finally, even if all injurers and victims wereincluded in a Sindell action, each injurer would not necessar-ily compensate his victims for the harm that he caused.Rather, the effect of the market share theory is that a classpays a class, with each individual defendant paying a portionof each individual plaintiff's damages. Every defendant paysfor some, but not all, of the harm he actually did and forsome harm done by others in the group. The IR principle istherefore violated for every defendant!

The court's disclaimer does not alter the difficulty of de-fending Sindell under the IR principle. On the other hand,Sindell can be readily understood as an application of a prin-ciple of group responsibility, whereby a plaintiff who is un-able to point to a responsible individual can point instead to

36. See id., 607 P.2d at 937-38, 163 Cal. Rptr. at 145-46.37. See, e.g., id. at 618-19, 607 P.2d at 940, 163 Cal. Rptr. at 148 (Richardson,

J., dissenting); Note, DES: Judicial Interest Balancing and Innovation, 22 B.C.L. REV.

747, 776-78 (1981).

14851986]

UCLA LA W RE VIEW

a responsible group and hold individual members of thegroup liable for each other's actions.

C. Scholarship: Group Justice for the "Indeterminate Plaintiff"

Following the DES cases, scholars began examining theimplications and possible extensions of those decisions toother areas where the traditional causation requirement wasdifficult or impossible to satisfy, particularly in so-called"toxic torts" or mass exposure cases. 38 In these cases, theproblem was one or both of the following: (1) the source ofa toxic agent that damaged the plaintiff was, as in the DEScases, only generally but not specifically identifiable; (2) atoxic agent from a specifically identifiable source was knownto have harmed some individuals within an identifiablegroup, but the specific victims within that group could notbe identified. One commentator dubbed the latter case theproblem of the "indeterminate plaintiff," the mirror of theDES "indeterminate defendant" problem. 39

The inherent difficulty of satisfying the traditional cau-sation requirement in these cases is illustrated by the follow-ing example. A leak by a chemical producer exposes 1000individuals to a toxic agent known to cause a particular formof cancer. This form of cancer also occurs at a certain"background level" in the population because of other un-known environmental sources. Following the leak, thenumber of cancer cases in the exposed population rises wellabove the background level. Although it is statistically cer-tain that some of the postexposure cases were caused by theleak, no one can determine which cases are attributable tothe leak. If the traditional causation rule is followed and sta-tistical evidence is relied upon,40 a greater-than-one hun-dred percent postexposure increase in cancer incidence willresult in every cancer victim recovering one hundred per-

38. See, e.g., Delgado, supra note 12; Robinson, supra note 12; Note, Epidemio-logic Proof of Probability: Implementing the Proportional' Recovery Approach in Toxic Expo-sure Torts, 89 DICK. L. REV. 233 (1984) [hereinafter cited as Note, EpidemiologicProof]; Note, The Inapplicability of Traditional Tort Analysis to Environmental Risks: TheExample of Toxic Waste Pollution Victim Compensation, 35 STAN. L. REV. 575 (1983)[hereinafter cited as Note, Inapplicability of Traditional Tort Analysis]; Note, Tort Ac-tions for Cancer: Deterrence, Compensation, and Environmental Carcinogenesis, 90 YALELJ. 840 (198 1) [hereinafter cited as Note, Tort Actions for Cancer].

39. See Delgado, supra note 12, at 882-83.40. See infra notes 64-65, 127-34 and accompanying text.

1486 [Vol. 33:1473

BETWEEN TWO WORLDS

cent of his damages from the toxic leaker, even though manyhave acquired cancer for reasons unrelated to ihe leak.4

1

This result would violate the IR principle because the de-fendant would pay for injuries he did not cause. However, ifthe increase in cancer incidence is one hundred percent orless, or if statistical evidence alone is not considered suffi-cient, then no victim will recover, although many almost cer-tainly contracted cancer because of the leak.

The solution to this conundrum, according to Profes-sors Rosenberg and Delgado, who write separately but reachquite similar conclusions, is that courts must regard the situ-ation as one involving harm done by a class or an individualto a class. Courts then must modify the causation require-ment accordingly. 42 The implicit corollary to this solution isthat the underlying principle of responsibility must shiftfrom individual to group responsibility.

Delgado's proposal would require an indeterminateplaintiff to proceed against the defendant through a class ac-tion. Plaintiff's causation requirement would be limited to ashowing that defendant's action was the only possiblehuman cause of the type of injury suffered. Once plaintiffmet this burden, presumably by epidemiologic evidence, de-fendant would be presumed responsible for the injuries of avictim group corresponding in number to the excess of post-exposure injuries over the preexposure level. The totaldamages would be paid to the entire plaintiff class of ex-posed individuals and distributed according to various possi-ble schemes. 43 Delgado acknowledges, but does not discussin depth, the possibility of an indeterminate plaintiff varia-tion of the Sindell situation, in which a defendant group or

41. The incidence must increase by more than 100% to permit any recovery.If the statistical evidence of the increased incidence of disease is the sole evidenceof causation, a 100% or less increase in cancer incidence following the leak shows,for any individual victim, that there is a 50% or less probability that his cancer wascaused by the leak. There is an equal or greater probability that it resulted fromanother source. Under the traditional rule requiring proof by a preponderance ofthe evidence, this would be insufficient to prove causation, and no one could re-cover. Conversely, if the incidence increases by more than 100%, every victim canmeet the burden and recover because for each victim there is a greater-than-50%chance that his cancer was caused by the leak.

42. See Rosenberg, supra note 3, at 855, 884-85, 905-07; Delgado, supra note12, at 899-901.

43. Delgado, supra note 12, at 899-900, 901 & n.97. The method of calculat-ing total damages is not entirely clear. See id. at 901 n.96.

1986] 1487

UCLA LA W RE VIEW

class would pay the plaintiff group, each individual defend-ant paying according to his market share.44 Rosenberg'sproposal is very similar, although he frames it in terms ofimposing on the defendant liability "proportional" to theexcess risk that he created. 45

Both Rosenberg's and Delgado's proposals rely on theconception of group responsibility or group justice. Del-gado notes that in these types of cases, "a class is in a senseinjured." 4 6 He recognizes that a class award "undercom-pensates some victims while overcompensating others, a re-sult some class members will see as unfair," but he impliesthat fairness to the class as a whole is a legitimate alterna-tive.47 Rosenberg is even more forthright in facing the issueof plaintiffs' windfalls and group fairness:

[O]ne could.., logically describe the defendant's duty inaggregative terms as a duty extending from the defen-dant to a class-the exposed population. Such a viewsuggests that the defendant's wrongdoing inflicts loss onthe exposed population as a whole .... The proportion-ality rule simply holds the defendant liable for.., the...losses it has caused in the "body" of the exposed popula-tion .... the population as a whole gains no windfall ....Essentially, the aggregative conception envisions thatcourts will assess liability on behalf of the entire exposedpopulation .... 48

In effect, Rosenberg and Delgado shift the focus of thesearch for an alternative to the IR principle from the sourceto the target of responsibility. Just as the special causationrules imply that responsibility can rest on groups of individu-als, 4 9 so also responsibility can be owed to groups of individu-als. In other words, a group responsibility principle carries

44. See id. at 907-08.45. See Rosenberg, supra note 3, at 859. Under Rosenberg's formulation,

which also requires a class action, plaintiff's causation requirement would be toestablish-by epidemiologic and, when appropriate, by market share evidence-the "probability of causation assigned to the excess disease risk in the exposedpopulation, regardless whether that probability fell above or below the 50%threshold .... " Id. at 859. Defendant(s) would be liable for a portion of theaggregate injuries of the entire plaintiff class (the exposed population in this ex-ample) proportionate to this excess risk figure and market share.

46., Delgado, supra note 12, at 888.47. Id. at 892. This alternative is even more compelling because the result

under traditional rules is unfair to all-no recovery for anyone.48. Rosenberg, supra note 3, at 884-85; see also Note, Epidemiologic Proof supra

note 38, at 256-58; Note, Tort Actions for Cancer, supra note 38, at 852-55.49. See supra notes 22-37 and accompanying text.

1488 [Vol. 33:1473

BETWEEN TWO WORLDS

implications for victims as well as injurers. 50 Under the IRprinciple, a victim's claim can never rest on any basis otherthan injuries he has suffered individually. Under group re-sponsibility, a victim's claim can be based on injuries suf-fered by a group of which he was a member. Thus, thecorollary of group responsibility is group rights. 5'

While Rosenberg and Delgado do not articulate theseramifications in the shift from individual to group responsi-bility, the implications are inherent in their proposals. Theirfocus on the victim significantly widens the implications ofthe group responsibility principle. Not only can a plaintiffunable to identify an individual injurer point to an identifi-able group of possible injurers and hold individual membersof that group responsible; but a plaintiff unable to identifyhimself individually as a victim can point to an identifiableinjured group of which he is a member, hold the injurer re-sponsible to this group, and share in the compensation.From both directions, individual responsibility shifts togroup responsibility.

D. Scholarship: "Individual Justice" Based on Probability ofCausation and Risk Contribution

Another group of commentators seems to suggest thatthe problem of causation in the indeterminate plaintiff ordefendant case can be handled in a manner consistent withthe IR principle. 52 All of these scholars contemplate individ-ual victims recovering from individual injurers, and all relyto some degree on the use of statistical evidence to establish

50. Under the IR principle, an injurer is responsible only for harm that hedoes and only to victims he personally affects. Under group responsibility, he isresponsible for harms done by others in his group to victims he has not affectedpersonally. Rosenberg and Delgado imply that he is responsible even for harmsdone by others not in his group, at least when their victims are indistinguishablefrom his, i.e., are members of his victim's group. In other words, by affectingsome victims, an injurer becomes responsible for others in the victims' group.The point may also be framed from the victims' perspective as in the text.

51. See Weale, supra note 15. From a different perspective, group responsibil-ity places responsibility on one victim for other victims. When we are in a com-mon group, I cannot single out my injury in disregard of yours. Instead, Irecognize that a response to my situation must include.a response to yours as well;hence, our claim is one of responsibility owed to a group, or a claim of groupright.

52. See generally King, supra note 12; Robinson, supra note 12; Note, Epidemio-logic Proof supra note 38; Note, Tort Actions for Cancer, supra note 38.

1986] 1489

UCLA LA W REVIEW

causation and on the device of awarding recovery for anamount other than the victim's full actual damages. Despiteappearances, however, the underlying basis of these propos-als is not individual responsibility but group responsibility.5 3

There are two basic approaches taken by this group ofscholars. The first would impose liability based on risk con-tribution or probability of causation. The second approachwould hold defendants liable for loss of expected value.Under the first approach, suggested most notably by Profes-sor Robinson, the individual plaintiff who encounters diffi-culty proving traditional causation because of eitherdefendant or plaintiff indeterminacy could recover from anyindividual defendant under "a rule that imposes liability forthe creation of a risk and apportions liability according tothe magnitude of that risk." 54 That is, the court wouldmerely require the plaintiff to establish a 'statistical figure re-flecting the "probability of causation" by a particular de-fendant. The plaintiff then would recover a portion of hisdamages from that defendant. The amount recoveredwould be proportional to the "probability of causation" fig-ure, whether the probability was fifty-one percent or lowerand whether it was established by purely statistical evidenceor otherwise. Robinson's "probabilistic approach to causa-tion" 55 can encompass both the problem of the indetermi-nate defendant and that of the indeterminate plaintiff.Robinson's hypothetical examples represent a combinationof the two. 56

The second approach is best exemplified by the writingsof Professor King.57 King suggests that courts should per-mit the individual toxic exposure victim, for example, to re-cover against an individual defendant or defendant class for

53. The reasons for the discrepancy between substance and appearance inthese proposals will be discussed later. See infra text accompanying notes 136-48.

54. Robinson, supra note 12, at 749.55. Id. at 766.56. See id. at 750-51, 758, 762. Other writers, including Rosenberg, have

stressed the potential usefulness of a "proportional liability" rule in the indeter-minate plaintiff situation. See Rosenberg, supra note 3, at 858-59. However, whileRosenberg envisions a class action and class remedy, id. at 905-24, others followRobinson and adopt proportional liability for probability of causation as a rule forindividual victim-injurer suits. See Note, Epidemiologic Proof supra note 38, at249-55; Note, Inapplicability of Traditional Tort Analysis, supra note 38, at 612-18;Note, Tort Actions for Cancer, supra note 38, at 855-59.

57. See King, supra note 12.

1490 [Vol. 33:1473

BETWEEN TWO WORLDS

the present value of the increased risk that the victim bearsbecause of exposure. 58 This ordinarily will be measured byreduced life or health expectancy in statistical terms. Theexposure victim suffers the immediate loss of some preexist-ing chance of a healthy and extended life. He is thus enti-tled to recover this loss immediately. If he suffers no illeffects from the exposure he still keeps the award, but if hesuffers severe effects he receives no further compensation.The parallel to the individual plaintiff windfall-shortfallproblem in Delgado's class remedy is plain, 59 but the form ofthe remedy here retains an individual character.

Both Robinson's risk contribution theory and King'slost value theory give the impression that liability in thesecases can be imposed on the basis of injurers' individual re-sponsibility for risk-creation or destruction of value. Never-theless, while proposals like Robinson's and King's seem tomaintain the IR principle, they actually involve groupresponsibility.

The underlying logic of risk contribution liability is thatover time, an injurer will pay to a series or group of individ-ual victims an amount roughly equal to the aggregate dam-age he caused to some of the victims in that group. This isprecisely the kind of "responsibility to a group" phenome-non involved in Delgado's indeterminate plaintiff propo-sal. 60 On the other hand, if the case involves indeterminatedefendants, as in Robinson's hypothetical case of a cancercaused by one of three alternative sources, 61 risk contribu-tion liability in effect ascribes responsibility to a group of in-

58. See id. at 1370-73, 1383-84; see also Note, Epidemiologic Proof supra note 38,at 255. King draws an analogy to the patient who suffers paralysis due to abotched operation, but whose preexisting condition made the paralysis likely evenafter a perfect operation. Such a patient is allowed by some courts to recover forthe "destruction of the chance" of recovery. See King, supra note 12, at 1363-70;see also Hicks v. United States, 368 F.2d 626 (4th Cir. 1966); Kimball v. Scars, 59A.D.2d 984, 399 N.Y.S.2d 350 (1977); Kallenberg v. Beth Israel Hosp., 45 A.D.2d177, 357 N.Y.S.2d 508 (1974), aff'd, 31 N.Y.2d 719, 337 N.E.2d 128, 374N.Y.S.2d 615 (1975). The opinions in the foregoing cases are vague as to whetherrecovery in these cases is limited in proportion to the original probability of sur-vival, as King's approach would suggest. However, a concurring opinion in aWashington Supreme Court case explicitly cited King and argued for a propor-tional recovery rule in such cases. Herskovits v. Group Health Coop., 99 Wash.2d 609, 630-36, 664 P.2d 474, 485-87 (1983) (Pearson, J., concurring).

59. See supra text accompanying notes 46-48.60. See supra text accompanying notes 48-51.61. See Robinson, supra note 12, at 750, 762.

1986] 1491

UCLA LA W REVIEW [Vol. 33:1473

dividuals whose activities tend to cause similar injuries. Inevery individual case, risk contribution liability would holdeach of the multiple defendants partially responsible forwhat may be another's actions. The logic is that the totaldamages paid by each defendant over a number of individ-ual cases will approximate the aggregate harm he inflicted.Nevertheless, this is the logic of "responsibility on a group"reflected in the Sindell decision;62 it is not the logic of the IRprinciple. The risk contribution theory makes sense only interms of a group responsibility principle. 63

The lost value theory depends equally upon a group re-sponsibility premise. Moreover, an analysis of this theoryhelps to understand better the group premise of the riskcontribution theory. The essence of the lost value approachis its reliance upon probabilistic or statistical evidence to as-sess both the victim's preinjury "value" and the impact ofthe injurious exposure on that value.64 Statistical evidenceessentially calls for inferences about the individual basedupon information about some group of which he is a mem-ber.65 This is true whether the statistical argument is ap-

62. See supra text accompanying notes 36-37.63. Robinson notes that in adopting his theory, he rejects Cardozo's famous

injunction against imposing liability for "negligence in the air." Robinson, supranote 12, at 738-39. In other words, he believes that the defendant need not actu-ally harm the individual plaintiff to justify imposing liability on the defendant.Liability should lie for negligence, or risk imposition, alone. Therefore, underRobinson's theory, a plaintiff exposed to injury by two tortfeasors, but actuallyinjured by only one of them, would recover from the noninjurer as well as theinjurer. If the noninjurer's probabilistic contribution to risk is greater, plaintiffwould recover more from him than from the known actual injurer. See id. at 739,754-58. There could hardly be a clearer violation of the IR principle, from theperspectives of both the injuring and the noninjuring defendant. (This is not sim-ply a version of the rule of concentrating responsibility on a superceding interven-ing cause, because all proximate cause discussions presume that every proximatecause is also an actual cause.) From the victim's perspective, risk contributionliability would mean that a victim would never recover from any individual de-fendant the actual damage done to him by that defendant. Each individual casewould involve a shortfall or windfall to the plaintiff vis-A-vis the particular defend-ant. Again, the IR principle would be violated.

64. The victim's preexisting value is established by reference to actuarial evi-dence. This evidence reflects expected life, health, and productivity values for agroup of individuals sharing certain characteristics that the victim possessed priorto the incident. Likewise, the impact of the injury is established by epidemiologicevidence of the expected impact of the risk exposure on a "population" sharingcertain characteristics, in which the victim is included. See King, supra note 12, at1382-85.

65. See Dore, A Commentary on the Use of Epidemiologic Evidence in Demonstrating

1492

BETWEEN TWO WORLDS

plied to the injurer's conduct and its consequences or to thevictim's injury and its source. In effect, if an individual de-fendant is held liable to an individual victim for "lost value,"it is because the defendant is considered responsible to agroup, of which the plaintiff is a member. If every individualin that group recovered for lost value, the defendant wouldultimately pay an amount equal to the aggregate harm doneto the group, although each individual member would eitherreap a windfall or suffer a shortfall. This is identical, then, tothe "group justice" of Delgado's proposal, despite its indi-vidualistic appearance.

The more general point-is that any theory that relies pri-marily upon statistical evidence necessarily accepts argu-ments about the individual based on his group affiliation ormembership. These arguments are acceptable and evennormal under a group responsibility principle. They arecompletely antithetical to the IR principle. The central roleof probabilistic and statistical evidence of causation both inKing's lost value theory and in Robinson's risk contributiontheory implies that each theory rests upon and representsgroup responsibility. Both these theories, regardless oftheir emphasis on actions by individual victims against indi-vidual injurers-with the apparent implication of loyalty tothe IR principle-are only coherent in terms of the groupresponsibility principle.

E. Recent Decisions. Applications of Theory

A number of recent decisions have demonstrated thatthe theories just described are having an impact on thecourts. These decisions extend group responsibility evenfurther than Bichler, Summers, Sindell, and other cases whicharticulated "special rules of causation" that implicitlyadopted group responsibility. In Collins v. Eli Lilly & Co.,"" aDES case, the Wisconsin Supreme Court concluded thatnone of the special causation rules were applicable and thenarticulated a variant theory involving group responsibility toan even greater degree. Explicitly citing Professor Robin-son's article, the court adopted a more radical version of his

Cause-in-Fact, 7 HARV. ENVTL. L. REv. 429, 431, 436-37 (1983); Note, EpidemiologicProof supra note 38, at 237, 239-40.

66. 116 Wis. 2d 166, 342 N.W.2d 37, cert. denied, 469 U.S. 826 (1984).

1986] 1493

1494 UCLA LAW REVIEW [Vol. 33:1473

risk contribution theory. 67 According to the court, the basisfor liability was that "[e]ach defendant contributed to therisk of injury to the public and, consequently, the risk of in-jury to individual plaintiffs . "...",68 The court therefore dis-pensed with the Sindell requirement of joining defendantswho represent a substantial share of the market and insteadheld that "plaintiff need commence suit against only one de-fendant .... "69 Under Collins, as Robinson had proposed, aplaintiff can thus sue any single party for "possible causa-tion" of injury. 70

The Collins court went even further. After supposedlyadopting the risk contribution theory, the court then heldthat "the plaintiff may recover all damages from the one de-fendant." 7' This means that Collins casts one hundred per-cent responsibility in every individual case on any "possiblecause"!7 2 This goes further than the original Summers rule,which imposes one hundred percent liability on each possi-ble cause only when all are sued. 73 It also goes beyond theproportional liability of Sindell and the Robinson proposal.74

In effect, the Collins court decided that even if the plaintiffcan point to only one member of an injurer group, that indi-vidual can be held fully responsible for the actions of every

67. See id. at 191 n.10, 342 N.W.2d at 49 n.10.

68. Id. at 191, 342 N.W.2d at 49.

69. Id. at 193, 342 N.W.2d at 50.

70. See Robinson, supra note 12, at 752-53; supra text accompanying notes54-56.

71. 116 Wis. 2d at 194, 342 N.W.2d at 50.

72. Either the plaintiff or defendant may be able to join other possible causalactors. If they do so, a comparative negligence scheme would proportionally limitthe liability of each defendant. However, if other possible causal actors cannot bejoined, the single possible cause must pay for all the plaintiff's damages.

73. See Summers v. Tice, 33 Cal. 2d 80, 86-88, 199 P.2d 1, 4-5 (1948); see alsoSindell v. Abbot Laboratories, 26 Cal. 3d 588, 603, 607 P.2d 924, 930-31, 163Cal. Rptr. 132, 138-39, cert. denied, 449 U.S. 912 (1980); RESTATEMENT (SECOND)OF TORTS § 433B(3) (1965); W. PROSSER & P. KEETON, supra note 1, § 41, at

270-71. Under a Summers rule, the individual bore some risk of shouldering adisproportionate share if codefendants were insolvent. Nevertheless, the individ-ual could generally count on some contribution and was shielded completely ifany potential codefendant (possible cause) was unavailable for the suit.

74. See Sindell, 26 Cal. 3d 612-14, 607 P.2d at 937-38, 163 Cal. Rptr. at145-46; Robinson, supra note 12, at 739, 749. Under Sindell and the Robinsonproposal, the risk of vastly disproportionate liability is even less than under Sum-mers, because liability is always limited in proportion to market share orprobability of causation. See supra notes 34-36, 54-56 and accompanying text.

BETWEEN TWO WORLDS

other absent member of his group. 75

A second example of the effects of recent scholarship isthe decision by the Federal District Court of Utah in Allen v.United States. 76 In this case, radiation generated by nucleartests conducted in the Nevada desert allegedly caused inju-ries to residents of the area. The court was presented with asituation in which the primary evidence of causation was thestatistical correlation between exposure to radiation and dis-ease incidence. None of the injuries were "specifically trace-able to the asserted cause on an injury-by-injury basis." 77

Nevertheless, the court cited "indeterminate plaintiff" pro-posals by Delgado and others, as well as theories supportingthe sufficiency of probabilistic "causal linkages" to establishcausation. 78 Relying on these theoretical bases-all pre-mised upon group responsibility-the court held that therequisite element of causation had been satisfied: "Where adefendant . . . negligently creates a radiological hazardwhich puts an identifiable population group at increasedrisk, and a member of that group ... develops a ... condi-tion which is consistent with having been caused by the haz-ard ... a fact finder may reasonably conclude that the hazardcaused the condition .. .

Allen thus appears to accept the sufficiency of probabilis-tic proof of causation, with its implicit group responsibility

75. Subsequent to Collins, the Washington Supreme Court allowed 100% re-covery from one possible causal actor in a DES case. See Martin v. Abbot Labora-tories, 102 Wash. 2d. 581, 604, 689 P.2d 368, 382 (1984). However, theWashington court allowed the single defendant to escape full liability by provingspecific market share, whereupon plaintiff's recovery would be limited to that por-tion of damages. Id. at 605-06, 689 P.2d at 383. The Collins court allowed thesingle defendant no similar opportunity; it insisted that the plaintiff be assured offull recovery from some source.

76. 588 F. Supp. 247 (D. Utah 1984).77. Id. at 406.78. Id. at 407, 414-16 (citing Calabresi, Concerning Cause and the Law of Torts: An

Essay for Harry Kalven, Jr., 43 U. CHI. L. REV. 69, 72 (1975); Delgado, supra note 12;Thode, Tort Analysis: Duty-Risk vs. Proximate Cause and the Rational Allocation of Func-tions Between Judge and Jury, 1977 UTAH L. REV. 1, 5-6; Note, Inapplicability of Tradi-tional Tort Analysis, supra note 38).

79. Allen, 588 F. Supp. at 415 (emphasis in original); see Ball, The Problems andProspects of Fashioning a Remedy for Radiation Injury Plaintiffs in Federal District Court:Examining Allen v. United States, 1985 UTAH L. REV. 267. Ball indicates that thegovernment is appealing the decision, and gives his view that, in the long run,legislation may be a better and more effective remedy. See id. at 314, 315-24. Butsee infra note 134, discussing the lack of legislative action on compensationschemes.

19861 1495

UCLA LAW REVIEW

premise. 80 However, the court does not limit the victim's re-covery in proportion to the probability proven, as Robinsonand King suggest.8 ' Instead, the court allows a recovery ofone hundred percent of damages to each individual plain-tiff.82 This rule provides no averaging effect over a series ofindividual cases. Consequently, the defendant ultimatelywill be held responsible for far more than the aggregatedamage he caused. This result attenuates the IR principleeven further.

Finally, in a decision that reflected many of the theoreti-cal proposals previously discussed, the Federal DistrictCourt for the Eastern District of New York recently ap-proved a settlement in In re Agent Orange, 8 3 the complex liti-gation over the effects of the defoliant "Agent Orange" onU.S. veterans who served in Vietnam. Although describedas "reasonable under the law,''84 the agreement reached inthis case presented the unprecedented solution of awardingrecovery to a class of plaintiffs and "distributing the dam-ages charged to all defendants as a group among all classmembers on a pro rata basis." 85 The court acknowledgedthat "it is doubtful whether the legal system is ready to em-ploy this device except, perhaps, as part of an overall settle-ment plan voluntarily entered into by the parties."8 6 Thecourt viewed the case as presenting a combination of the in-determinate defendant and indeterminate plaintiffproblems, together with the issues of the sufficiency of prob-abilistic evidence and the appropriateness of nonindividual-'ized treatment of plaintiffs' claims.87 In effect, the casecombined all the elements of the different hypotheticalsused by the different theoretical proposals to explain and ar-gue for group responsibility. To impose liability in this situ-ation would involve group responsibility in both senses:responsibility imposed on a group and responsibility owedto a group. Despite the lack of firm precedent, the court

80. However, see infra text accompanying notes 135-40 for a different level ofanalysis.

81. See supra text accompanying notes 54-59.82. Allen, 588 F. Supp. at 446-47.83. In re Agent Orange Prods. Liab. Litig., 597 F. Supp. 740 (E.D.N.Y. 1984).84. See id. at 749.85. Id. at 748.86. See id.87. Id. at 819-43.

1496 [Vol. 33:1473

1986] BETWEEN TWO WORLDS 1497

stated: "We should not unduly restrict development of legaltheory and practice .. by dismissing a class action such asthe one now before us ... ."88 This statement suggests thatthe Agent Orange decision may become a precedent for futuretoxic torts decisions that recognize and apply the theory ofgroup responsibility to both defendants and plaintiffs. 89

88. Id. at 749, 842-43.89. Another frequent response to the problem presented by the traditional

causation requirement, particularly in situations in which plaintiff or defendantidentification poses substantial difficulties, is to advocate abandoning not only theidentification requirement, but the torts adjudication process as a whole. Such aresponse is particularly appealing in relation to cases like Collins and Agent Orange,in which the attempt to find a remedy within the torts system threatens to stretchtort doctrines beyond the breaking point. This has led various authors to proposethe creation of administrative mechanisms outside the torts system to compensatevictims of injuries such as those resulting from toxic exposure. See, e.g., Downey &Gulley, Theories of Recovery for DES Damage. Is Tort Liability the Answer, 4 J. LEGALMED. 167, 196-200 (1983); Ginsberg & Weiss, Common Law Liability for Toxic Torts:A Phantom Remedy, 9 HOFSTRA L. REV. 859, 928-40 (1981); Schwartz & Mahshigian,Failure to Identify the Defendant in Tort Law: Towards a Legislative Solution, 73 CALIF. L.REV. 941, 964-74 (1985); Note, Sindell v. Abbott Laboratories: Is Market ShareLiability the Best Remedy to the DES Controversy?, 18 CAL. W.L. REV. 143, 172-77(1981) [hereinafter cited as Note, Best Remedy]; Note, Torts-Market Share Liability-The California Roulette of Causation Eliminating the Identification Requirement-Sindell v.Abbott Laboratories, 11 SETON HALL L. REV. 610, 627-28 (1981) [hereinaftercited as Note, Eliminating the Identification Requirement]; Note, Inapplicability of Tradi-tional Tort Analysis, supra note 38, at 589-91; Note, Industry-Wide Liability, 13 SUF-FOLK U.L. REV. 980, 1015-22 (1979).

While these proposals change the context of the debate over group responsi-bility, they do not avoid the issue. In fact, these proposals themselves form part ofthe shift from individual to group responsibility: the proposed compensationschemes almost always rely on a group responsibility premise. The common formof these proposals suggests the creation of a compensation fund, usually by levy-ing some form of tax on producers whose processes use or create toxic substancesthat tend to produce certain kinds of injuries. See, e.g., Ginsberg & Weiss, supra, at928-40; Note, Best Remedy, supra, at 172-77. In general, the charges are ex-antelevies unrelated to the involvement of the individual producer in specific cases ofinjury. (However, the fund may be supplemented by amounts recovered fromspecific injurers who are proven responsible in actions brought by the fund itself.See Note, Inapplicability of Traditional Tort Analysis, supra note 38, at 590-91.) Vic-tims can recover from the fund by showing that they suffered a type of injuryassociated with toxic exposure after an incident of possible exposure. Typically,damages are scheduled or otherwise standardized. In some versions, recovery islimited in proportion to the probability of causation the plaintiff can establish.

Thus, the typical proposed compensation scheme involves group responsibil-ity from both the defendant and the plaintiff perspectives. By rejecting the possi-bility of funding such a scheme from general tax revenues, most proposals clearlyimply that, lacking a responsible individual, it is legitimate and preferable to affixresponsibility upon an identifiable and limited group, rather than diffuse it amongthe population as a whole. See G. CALABRESI, supra note 3, at 64-66, 255-63. Thevictim points to a (preestablished) group of individuals involved in the kind ofactivity that injured him, and each individual in this group, by his ex-ante pay-

UCLA LA W REVIEW

F. "Mass Torts" vs. "Sporadic Accidents".The Wider Implications

One response to the foregoing discussion is that whilethe decisions and theories analyzed may involve group re-sponsibility concepts, they are all limited to the narrow areaof "mass torts" committed by business enterprises produc-ing undifferentiated products. Therefore, the group respon-sibility principle they adopt will simply not apply to the vastmajority of torts cases, which will continue to involve "spo-radic" accidents committed by individuals and enterprisesdealing in identifiable actions and products. 90 It can thus beargued that the shift from individual to group responsibility,however well-documented in the mass torts context, carrieslittle significance for the average citizen, who is unlikely tobe exposed to group responsibility because he will rarely beinvolved in the kinds of activities that cause mass torts.

This argument seriously understates the significance ofthe shift from individual to group responsibility. First, theshift encompasses not only the injurer's but also the victim'sperspective. 91 Cases that hold injurers responsible togroups of individuals, only some of whose members are ac-tual victims, manifest the group responsibility principle.The average citizen may well be a victim in a mass tort inci-dent; and the adoption of group responsibility for mass tortswill seriously affect the individual as victim. Specifically, itwill expose him to the real possibility of receiving less thanfull compensation while his neighbor receives a windfall for

ments, is held responsible for some share of an injury that may have been causedby another. From the victim's perspective, he receives some compensation fromthe fund, even though his injuries may have been caused by a source unrelated toany of the fund's contributors. However, he receives a standardized award thatmay be less than the actual injuries he suffered, even when they were caused by afund-member's actions. These results represent group responsibility, becausethey compensate victims in the aggregate accurately, i.e., to the full extent of ag-gregate fund-related injuries. However, individual victims receive shortfalls orwindfalls. Compensation schemes proposed as alternatives to the tort system, in

part because of the causation requirement, thus represent and reflect the shiftfrom individual to group responsibility. See i'fra note 106 for a discussion of thedistinction between these schemes and "social insurance" schemes that involvesocietal or collective rather than group responsibility.

90. The distinction between mass torts and sporadic accidents is suggested byRosenberg, among others. See Rosenberg, supra note 3, at 854-55, 858-59 passim:see aLw Williams, Mass Tort Class Actions: Going, Going, Gone?, 98 F.R.D. 323 (1983).

91. See supra notes 38-51 and accompanying text.

[Vol. 33:14731498

BETWEEN TWO WORLDS

an unrelated condition. The dissatisfaction expressed by in-dividual victim group members with class awards or settle-ments in mass tort cases attests to the serious consequencesof group responsibility for individual citizens in this area. 92

Second, an example will demonstrate that the distinc-tion between mass torts and sporadic accidents is ultimatelyillusory with respect to the application of group responsibil-ity. Suppose that an individual drove carelessly down astreet where a pedestrian stood on the curb waiting to cross.Just as the driver reached the point where the pedestrianstood, a careless tree-trimmer working directly over the pe-destrian's position dropped a tree limb. Simultaneously, alarge, unleashed dog ran past the pedestrian. At that mo-ment, something-the pedestrian could not afterward saywhat-caused the pedestrian to fall, and he suffered a seri-ous head injury. The dog disappeared, the tree-trimmerwent bankrupt, and the pedestrian sued 'the driver for onemillion dollars. Several people witnessed the occurrence:some testified that the tree limb knocked the pedestriandown, some that the dog tripped him, and some that thedriver actually hit him. No one was obviously lying and thephysical evidence of impact was ambiguous. The evidencewas truly inconclusive. It probably would be impossible forplaintiff in this case to establish causation under the tradi-tional rule requiring proof of causation by a preponderanceof the evidence. The individual driver, although careless,would be free of liability, and most of us would say rightlyso, because his individual responsibility for the harm couldnot sufficiently be established.

However, under Robinson's risk contribution theory,for example, any probability of causation, whether fifty-onepercent or less, would provide grounds for imposing pro-portional liability on the individual driver. 93 If the juryfound that there was a ten percent chance that the driveractually hit the plaintiff, this "possibility of causation" wouldresult in the imposition of $100,000 liability on the driver,who was by hypothesis almost surely innocent of inflictingthe harm.94 Robinson never states that his theory would ap-

92. See Note, Epidemiologic Proof supra note 38, at 247 & n.88.93. See Robinson, supra note 12, at 755-60, 764-67.94. Take another example of a slightly different kind. Suppose that a prop-

erty owner fails to light a stairway, adequately, and a visitor slips and falls, suffer-

1986] 1499

1500 UCLA LAWRE VIEW [Vol. 33:1473

ply only to mass torts. On the contrary, he clearly impliesthat his theory would alter the basis of tort liability acrossthe board. Many cases of causal ambiguity, like the preced-ing hypothetical, involve individuals as possible causes andusually result in findings of no liability under the traditionalrule. The sweeping alteration of this rule certainly wouldhave a major impact on individual citizens as injurers, andthis impact would ultimately stem from the use of a groupresponsibility principle to impose liability. The same can besaid of other proposals involving group responsibility. 95

ing a crippling injury. Suppose also that no evidence about the actual cause of thefall can be found, other than the fact of the fall and statistical data showing that incases of falls on unlighted stairs, the lack of light is a contributing cause of the fall88% of the time. Under a strict application of the traditional causation require-ment, with its IR premise, it might be impossible for plaintiff to prove causation inthis case. See infra notes 127-40 and accompanying text. It certainly would beimpossible if the statistical correlation were only 48%. In either case, even ifsome additional evidence of causation is available, it might be insufficient to con-vince a jury that the absence of light more likely than not was a necessary cause ofthe fall. The property owner would be found free of liability because his individ-ual responsibility for the harm could not be established.

In contrast, under Robinson's risk contribution theory, any probability ofcausation-whether 51% or less and whether established by statistical evidencealone or by more specific evidence-would provide grounds for imposing propor-tional liability on the individual property owner. Even if the victim admitted thatthe cause of her fall was that her heel broke a minute before the fall, the individualproperty owner would still be proportionately liable for risk contribution alone.See id. at 755-60, 764-67; see also Wolf v. Kaufmann, 227 A.D. 281, 237 N.Y.S. 550(1929) (dismissing complaint by decedent's heir because there was no eyewitnesstestimony or other specific evidence that the unlighted stairway caused the fall).But see Reynolds v. Texas & Pac. Ry., 37 La. Ann. 694 (1885) (affirming judgmentfor plaintiff in essentially identical circumstances, based solely on the court's per-ception of "the natural and ordinary course of events" in these situations, that is,a rough idea of statistical correlation).

95. Proposals urging greater reliance on statistical and probabilistic evidence,although often made in the context of mass torts and toxic torts, could applyequally to any sporadic accident case in which statistical evidence of causation isavailable. See, e.g., Note, Epidemiologic Proof supra note 38; Note, Inapplicability ofTraditional Tort Analysis, supra note 38; Note, Tort Actions for Cancer, supra note 38; seealso King, supra note 12. Similarly, under some proposed accidental injury com-pensation schemes, the hypothetical individual property owner discussed supranote 94, along with all other multilevel property owners, might be charged ayearly tax corresponding to his pro rata share of the aggregate harm expected tooccur during that year from darkened stairway falls. The tax would go to a fundfrom which stairway fall victims could recover. See Franklin, Replacing the NegligenceLotteiy: Compensation and Selective Reimbursement, 53 VA. L. REV. 774 (1967);O'Connell, Expanding No-Fault Beyond Auto Insurance: Some Proposals, 59 VA. L. REV.749 (1973); O'Connell, No-Fault Liability by Contract for Doctors, Manufacturers, Re-tailers, and Others, 1975 INs. LJ. 531; see also supra note 89.

In fact, compensation advocates have made such proposals for ordinary or

1986] BETWEEN TWO WORLDS 1501

The shift from individual to group responsibility in thecontext of the traditional tort doctrine of causation thereforemust be seen as a phenomenon of major significance andpotentially widespread impact. Actually, the shift has alsooccurred in other areas of tort doctrine 96 as well as in other

sporadic accidents as well as for mass torts. The proposals for sporadic accidentcompensation, including the original automobile accident plans,- actually predatethe toxic torts proposals. This suggests that the group responsibility principlebehind such schemes is by no means limited to mass torts committed by imper-sonal conglomerates. Rather, the principle of group responsibility can apply tothe whole panoply of tort situations and thus have far-reaching effects on the indi-vidual citizen, both as potential injurer and, as discussed above, as potentialvictim.

96. The doctrine of strict liability for defective products is one example of theshift from individual to group responsibility outside the causation area. One basisfor imposing strict liability on manufacturers is the product's failure to live up tothe reasonable consumer's expectations of safety. See Barker v. Lull Eng'g Co., 20Cal. 3d 413, 429-30, 573 P.2d 443, 454, 143 Cal. Rptr. 225, 236 (1978); RESTATE-

MENT (SECOND) OF TORTS § 402A and comments g & i (1965). In effect, the basisof liability is quasi-contractual: It arises from a breach of the warranty implied bydefendant's marketing the product. Generally, courts focus their inquiry in thesecases on what expectations the individual defendant's marketing conduct couldhave created in a reasonable consumer. However, some commentators have sug-gested that the proper focus of inquiry is what expectations were created amongconsumers as a class by the marketing conduct of producers as a class. If theproducer class created high expectations of safety, consumers should be able torecover when those expectations are not fulfilled, regardless of the individual pro-ducer's marketing behavior. The individual producer should be held responsiblefor the conduct of the group with which he is associated. See, e.g., Rintala, TheSupreme Court of California 1968-1969-Foreword: "Status" Concepts in the Law ofTorts, 58 CALIF. L. REV. 80, 85-94 (1970). Rintala suggests approvingly thatcourts actually follow this group responsibility approach in product liability casesand in other cases involving quasi-contractual tort liability. See id. at 95, 101,104-05, 126-29.

Other commentators have observed that traditional strict liability rules thatimpose liability without regard to individual fault on those who engage in abnor-mally dangerous activities, see RESTATEMENT (SECOND) OF TORTS §§ 519-520(1965), serve to place responsibility on individual actors because of the riskinessof the conduct of the group they belong to as a whole. From this perspective,these rules can be seen as applications of the group responsibility principle to thenegligence issue. In effect, one who is not personally negligent nevertheless maybe held responsible because the group of which he is a part created an unreasona-ble level of risk. See, e.g., Calabresi & Hirschoff, Toward a Test for Strict Liability inTorts, 81 YALE LJ. 1055, 1064-70 (1972); Fletcher, supra note 3, at 541-43, 549.The same commentators argue for the wider application of such group-basedrules of liability.

A third example of the shift to group responsibility in torts is perhaps mostsignificant, though less obvious. While superficially unrelated to the issue, thevast increase in the use of comparative fault has often meant the de facto applica-tion of group responsibility. The Uniform Comparative Fault Act provides that"[i]n determining the percentages of fault [and liability], the trier of fact shallconsider both the nature of the conduct of each party at fault and the extent of the

UCLA LA W REVIEW [Vol. 33:1473

areas of the law. 97 However, the foregoing examination ofthe causation requirement provides ample evidence of howthe shift is affecting court decisions and legal theories of ma-jor significance.

causal relation between the conduct and the damages claimed." UNIF. COMPARA-TIVE FAULT ACT § 2(b) (1977). Some commentators suggest that this statute, andthe many state statutes based on it, mandate a rule of "comparative causation,"similar in practice to Robinson's proposed rule of proportional liability, see supratext accompanying notes 52-65, by which any possible causal actor who is negli-gent can be held liable in proportion to the probability that he caused the acci-dent. See Twerski, The Many Faces of Misuse: An Inquiry Into the Emerging Doctrine ofComparative Causation, 29 MERCER L. REV. 403, 431-34 (1978). The KansasSupreme Court has observed that this view is not merely theoretical:

[T]he Kansas comparative negligence act is a multi purpose actwhich goes far beyond a basic comparison of the contributing negli-gence of each of the parties to the cause of an accident or injury.The act comprehensively provides machinery for drawing all possi-ble parties into a lawsuit to fully and finally litigate all issues andliability arising out of a single collision or occurrence, and apportionthe amount of total damages among those parties against whom neg-ligence is attributable in proportion to their degree of fault.

McCart v. Muir, 230 Kan. 618, 622, 641 P.2d 384, 389 (1982); see also Eurich v.Alkire, 224 Kan. 236, 237, 579 P.2d 1207, 1208 (1978). Moreover, according tothe Kansas statute, any possible causal actor "shall bejoined as an additional party."KAN. STAT. ANN. § 60-258a(c) (1983) (emphasis added). This implies that, asunder Robinson's proposed approach, any negligent possible causal actor is sub-ject to proportional liability. Because proportional liability involves a form ofgroup responsibility, see supra text accompanying notes 52-65, the increasing useof "comparative causation" under the rubric of comparative fault is a major, if notnecessarily obvious, part of the shift to group responsibility.

97. Outside of tort law, there are many examples of the shift to group respon-sibility. One is a trend in contract law, similar to that in tort law, see supra note 96,to impose quasi-contractual liability on individuals based on expectations createdby group conduct. See, e.g., Tobriner & Grodin, The Individual and the Public ServiceEnterprise in the New Industrial State, 55 CALIF. L. REV. 1247 (1967). Another is thegrowth in the use of class actions and group litigation as an alternative to individ-ualized adjudication of claims. See, e.g., Hazard, The Effect of the Class Action DeviceUpon the Substantive Law, 58 F.R.D. 307 (1973); McGovern, Management of Mul-tiparty Toxic Tort Litigation: Case Law and Trends Affecting Case Management, 19 FORUM

1, 5-17 (1983); Rosenberg, supra note 3, at 905-16; Williams, supra note 90.Generally, the shift to group responsibility outside of the torts causation area

has been neither obvious nor explicit, except in the class action movement. Prob-ably for this reason, only the group causation decisions have aroused great con-troversy and hostility, although the class action phenomenon has met its ownwave of opposition. See, e.g., McGovern, supra, at 11; Rosenberg, supra note 3, at908-09; Weinstein, Some Reflections on the "Abusiveness" of Class Actions, 58 F.R.D.299, 306 (1973); Williams, supra note 90, at 331-35; Note, Federal Mass Tort ClassActions: A Step Toward Equity and Efficiecy, 47 ALB. L. REV. 1180, 1183-87 (1983);Note, Class Actions and Mass Toxic Torts, 8 COLUM. J. ENVTL. L. 269, 269-73 (1982);Note, Class Actions in New York: Recovery for Personal Injury in Mass Torts Cases, 30SYRACUSE L. REV. 1187 (1979).

1502

BETWEEN TWO WORLDS

II. RESPONSES TO THE SHIFT: RESIGNATION AND REJECTION

The decisions and proposals discussed in Part I haveevoked hostile responses from the majority of courts andcommentators. These responses have taken two distinctforms. The first accepts the causation theories discussedabove, but only as an accommodation to the needs of practi-cal justice for innocent victims. I call this the "necessaryevil" response. The second type of response rejects theabove theories, in whole or part, as unacceptable challengesto the IR principle, which is viewed as fundamental and invi-olable. I call this the "unmitigated evil" response. Neitherresponse views the new causation theories, or the group re-sponsibility principle underlying them, as an essentially pos-itive and progressive phenomenon. In Part II, I illustratethese responses and then argue that they stem from thecourts' deep commitment to the IR principle. This commit-ment is so powerful because the IR principle is inextricablylinked to the theory of political structure that prevails in ourlegal system and in the larger society as well. I conclude thatthe new causation theories, in any form, are unlikely to findwide acceptance until we explicitly examine the group re-sponsibility principle and its underlying political-philosophi-cal premises to determine whether they are trulyinconsistent with the IR principle and its premises.

A. Resignation: Group Causation as a Necessary Evil

Even the Sindell court was sensitive to the inconsistencybetween the market share liability doctrine and the IR prin-ciple. 98 The court insisted that there were, however, "force-ful arguments in favor of holding that plaintiff has a cause ofaction."9 9 The very first of these reasons was the argumentof necessity:

Advances in science and technology create fungiblegoods . . .which cannot be traced to any specific pro-ducer. The response of the courts can be to adhere rig-idly to prior doctrine, denying recovery ... or to fashionremedies to meet these changing needs .... [S]ome ad-aptation of the rules of causation ... may be appropriate

98. See Sindell v. Abbott Laboratories, 26 Cal. 3d 588, 612-14, 607 P.2d 924,937-38, 163 Cal. Rptr. 132, 145-46, cert. denied, 449 U.S. 912 (1980); see also supratext accompanying notes 36-37.

99. Sindell, 26 Cal. 3d at 610, 607 P.2d at 936, 163 Cal. Rptr. at 144.

15031986]

UCLA LA W REVIEW

in these recurring circumstances. 100

In essence, the court recognized that the new causation rulesdo compromise the IR principle and that this result is indeedregrettable. However, this compromise is a necessary expe-dient to achieve "practical justice" by somehow compensat-ing innocent victims of producer negligence. Likewise, in arecent pro-plaintiff DES decision, Martin v. Abbott Laborato-ries, 10, the Washington Supreme Court explicitly acknowl-edged the trade-off between providing some compensationto victims and honoring the IR principle: "We arepresented with a conflict between the familiar principle thata tortfeasor may be held liable only for damages that it hascaused, and the sense of justice which urges that the victimsof this tragedy should not be denied compensation... ."102

Echoing a similar sentiment, the Wisconsin Supreme Courtruefully admitted in the Collins case that "some of the ...defendants may be innocent, but we accept this as the pricethe defendants, and perhaps ultimately society, must pay toprovide the plaintiff an adequate remedy . ... "103 Othercourts, in both DES and asbestos cases, have articulated thesame kind of justification.104 Thus, even the most support-ive courts acknowledge that the new rules weaken and un-dermine the IR principle, and they regard this as an evil.But they conclude, somewhat uncomfortably, that it is a nec-essary evil because innocent victims otherwise will be leftwithout a remedy.

A number of post-Sindell commentators have joined thecourts and justified the decision and its progeny as a neces-sary evil-a mixed blessing at best. 0 5 Some commentators

100. Id., 607 P.2d at 936, 163 Cal. Rptr. at 144.101. 102 Wash. 2d 581, 689 P.2d 368 (1984).102. Id. at 603, 689 P.2d at 381.103. Collins v. Eli Lilly'& Co., 116 Wis. 2d 166, 198, 342 N.W.2d 37, 52, cert.

denied, 469 U.S. 826 (1984).104. The Bichler Court, for example, justified its expanded concert of action

rule by reference to the "exigencies of trying a case in the rapidly developing area... of strict products liability ... [and] the exigencies of rapidly developing tech-nologies." Bichler v. Eli Lilly & Co., 79 A.D.2d 317, 326, 328, 436 N.Y.S.2d 625,630-31, 632 (1981), af'd, 55 N.Y.2d 571, 436 N.E.2d 182, 450 N.Y.S.2d 776(1982); see also Hardy v. Johns-Manville Sales Corp., 509 F. Supp. 1353, 1358(E.D. Tex. 1981), rev'd, 681 F.2d 834 (5th Cir. 1982); Copeland v. Celotex Corp.,447 So. 2d 908, 913 (Fla. Dist. Ct. App. 1984), rev'd in part and modified, 471 So. 2d533 (Fla. 1985).

105. One recent Note concludes that "an alternative liability theory . . .maycause traditional concepts and basic principles of tort law to be distorted or aban-

1504 [Vol. 33:1473

1986] BETWEEN TWO WORLDS 1505

add another note to the "necessary evil" argument. Recog-nizing that concern for victim compensation alone mightjus-tify a social insurance scheme as easily as special causationrules in tort law,' 0 6 they argue that increased deterrence is asecond mitigating factor that weighs in favor of using groupcausation rules in tort, despite the fundamental conflict be-tween these rules and the IR principle.t0 7 For example, Del-

doned. . . .[Nevertheless,] our traditional legal notions must evolve to keep upwith our progressing society." Note, Proving Causation in Toxic Torts Litigation, 11HOFSTRA L. REV. 1299, 1325 (1983); see also Note, DES and a Proposed Theory ofEnterprise Liability, 46 FORDHAM L. REV. 963, 1002, 1007 (1978).

106. See Note, supra note 3, at 329; Note, Manufacturers'Liability Based on a Mar-ket Share Theory: Sindell v. Abbott Laboratories, 16 TULSA L.J. 286, 315 (1980)[hereinafter cited as Note, Manufacturers' Liability]. For those concerned above allelse with compensating victims, the most effective measure to ensure compensa-tion is social insurance. See G. CALABRESI, supra note 3, at 46-47; Epstein, ProductsLiability: The Search for the Middle Ground, 56 N.C.L. REV. 643, 659 (1978). Such asystem establishes a compensation fund out of general taxes, which provides thelargest possible revenue base, and compensates all accident victims from thisfund, whatever the circumstances of their injury. See Henderson, The New ZealandAccident Compensation Reform, 48 U. CHI. L. REV. 781 (1981); Palmer, Accident Com-pensation in New Zealand: The First Two Years, 25 AM. J. CoMP. L. 1 (1977); see alsoinfra text accompanying notes 241, 271. This system is identical in form to privateloss insurance (also referred to as casualty, or first party, insurance) funded byindividual premiums, but a socialized revenue base and government administra-tion make universal compensation more of a reality under social insurance. Thus,social insurance usually means socialized loss insurance.

Pure loss insurance, whether private or socialized, should be distinguishedfrom private and public insurance mechanisms, which, while they compensate vic-tims, charge the cost of compensation neither to the victims nor the public atlarge, but to individuals or groups considered responsible for the injury. In the-ory, this result places a deterrent burden on injury-causing actors and groups. SeeG. CALABRESI, supra note 3, at 46. But it also narrows the resource base for victimcompensation. Therefore, if compensation is the supreme goal, any such schemewould be inferior to social insurance. The most obvious example of such "costimposition insurance" is private liability insurance. See infra note 123. However,most public "nonfault" injury compensation systems involve at least some mea-sure of cost imposition on specific individuals or groups. See supra note 89. In thecontext of this Article, the significance of the distinction between these types ofinsurance is that social insurance is a form of societal or collective responsibility,whereas cost imposition insurance is a form of group responsibility. Neither in-volves individual responsibility.

107. There is some irony in the argument that the special causation rulesshould be preferred because they increase deterrence, despite their negative im-pact on the IR principle. See infra notes 108-10 and accompanying text. The de-terrence argument is based essentially on the utilitarian view that the tort systemis a vehicle for minimizing social costs of accidental injury. This view is mostpersuasively argued by the law and economics theorists, including Calabresi andPosner, who share this general view even if they disagree about specific legalrules. See, e.g., G. CALABRESI, supra note 3, at 26-31; Calabresi, supra note 3, at656; Calabresi & Hirschoff, supra note 96, at 1057; Posner, supra note 3, at 31-34.

1506 UCLA LAW RE VIEW [Vol. 33:1473

gado argues that the greater exposure of defendants toliability under group causation rules "deter[s] the defend-ants from engaging in the liability generating practice" and"encourages them to investigate the way in which their ac-tions endanger others."' 10 8 Rosenberg agrees that groupcausation rules increase deterrence and produce both saferactivities and more safety research. 10 9 At the same time, heis candid about the negative effects on the IR principle, ac-knowledging that the group causation/class action approach

The economic-utilitarian view, as applied by Posner and Calabresi, typically ana-lyzes conduct in group terms rather than in individual terms. As one commenta-tor aptly states: "Calabresi stipulates that the problems of accident law must betreated in terms of activities rather than careless [individual] conduct. 'An activitymay properly be defined as the doing of something by an actuarial class, whichmay tend to do it carelessly.' " Englard, supra note 3, at 40 (quoting Calabresi, TheDecision for Accidents: An Approach to Nonfault Allocation of Costs, 78 HARV. L. REV.

713, 715-16 (1965)); see also G. CALABRESI, supra note 3, at 246-50, 255-59,282-85; Calabresi & Hirschoff, supra note 96, at 1067-70, 1075. According toEnglard, although Posner disagrees with Calabresi's desire to abandon the indi-vidual case adjudication system of tort law, Posner does believe that "[individual]litigants ... serve only as the appropriate means to achieve an overall economicaim. The law is not concerned with the plaintiff's quest for personal redress. Itviews him as a convenient tool ...." Englard, supra note 3, at 54. Posner himselfdeclares: "[T]he economist is not interested in the one question that concerns thevictim .. .: [W]ho should bear the cost of this accident? .. .The issue becomeswhat is a just and fair result for a class of activities . R. POSNER, ECONOMIC

ANALYSIS OF LAw 18-19 (2d ed. 1977).The economic interpretation of the deterrent goal of torts is therefore a fairly

clear application of group responsibility. That this theory should lend support togroup causation rules is hardly surprising. It is merely circular to argue thatgroup causation rules should be favored, despite damage to the IR principle, be-cause they further deterrence. This obscures the point that both group causationand the group-oriented economic premises of deterrence theory conflict with theIR principle and its individualistic premise. Properly understood, the deterrenceargument for the special causation rules only emphasizes the conflict with the IRprinciple. It cannot be offered, as it often is, to settle the conflict. Most courtshave at least intuitively grasped this point and have refused to adopt group causa-tion rules on deterrence grounds. See infra text accompanying notes 157-86.

108. Delgado, supra note 12, at 893-94.109. Rosenberg, supra note 3, at 877 ("[T]he loss of deterrence that results

from applying either version of the [traditional] preponderance rule in mass expo-sure cases is likely to be direct and very substantial."). A Note proposing adop-tion of a rule allowing' recovery based on statistical proof of the possibility ofcausation in cancer cases phrased the same idea more technically: "The proposalwould force producers [of carcinogens] to internalize the actual social costs oftheir activities, thereby providing an incentive for producers to reduce their emis-sions ... to socially optimal levels." Note, Tort Actions For Cancer, supra note 38, at859. Some of the courts adopting group causation rules also advert to the deter-rence rationale, although less explicitly than the commentators. See Sindell, 26 Cal.3d at 612-13, 607 P.2d at 936-37, 163 Cal. Rptr. at 144-45; Collins, 116 Wis. 2d at192-93, 342 N.W.2d at 49-50.

BETWEEN TWO WORLDS

he favors "seeks to achieve the benefits of deterrence by sac-rificing some of the benefits of individualized treatment ofclaims.", 10

The common theme is that the new rules are a mixedblessing-a necessary evil. They further victim compensa-tion and deterrence but inevitably weaken the IR principle.Thus, the shift from individual to group responsibility hasbeen met with ambivalence at best, even by the supportersof the new causation rules.

B. Resistance: Group Causation as an Unmitigated Evil

Among both courts and commentators, the two mostcommon responses to the whole spectrum of proposed mod-ifications of the traditional causation requirement have beeneither outright rejection or subtle resistance. The main rea-son behind these negative responses is an implicit or explicitcommitment to the IR principle, which is seen as fundamen-tally challenged by the new rules and proposals.

1. Rejection of Group Causation

Even the Sindell opinion contained a strong dissent thatwould have rejected market share liability. Justice Richard-son, the author of the dissenting opinion, succinctly articu-lated the minority's principal complaint: "The majority nowexpressly abandons the ... traditional requirement of somecausal connection between defendants' act and plaintiffs' in-jury... [rejecting] over 100 years of tort law which requiredthat before tort liability was imposed, a 'matching' of [an in-dividual] defendant's conduct and [an individual] plaintiff'sinjury was absolutely essential."' In other words, the courthad abandoned the IR principle along with the traditionalcausation requirement.

In Payton v. Abbott Laboratories, 112 a DES case in which

110. Rosenberg, supra note 3, at 907.111. Sindell v. Abbott Laboratories, 26 Cal. 3d 588, 615-16, 607 P.2d 924,

939, 163 Cal. Rptr. 132, 147 (Richardson,J., dissenting), cert. denied, 449 U.S. 912(1980).

112. 386 Mass. 540, 437 N.E.2d 171 (1982). Subsequently, the Federal DistrictCourt for Massachusetts adopted market share liability in a DES case, basing itsdecision in part on a comment by the Payton court that they might in the futureconsider adopting a form of market share liability different than that proposed byplaintiffs in Payton. McCormack v. Abbott Laboratories, 617 F. Supp. 1521, 1525(D. Mass. 1985). See also Payton, 386 Mass. at 574, 437 N.E.2d at 190.

1986] 1507

UCLA LA W REVIEW [Vol. 33:1473

the court rejected market share liability, the MassachusettsSupreme Court explicitly recognized the connection be-tween the IR principle and the traditional causation require-ment. "Identification of the party responsible for causinginjury to another is a longstanding prerequisite to a success-ful negligence action. This requirement . . . ensures thatwrongdoers are held liable only for the harm that they havecaused." 1 t 3 The court warned of the consequences ofadopting the market share theory: "[It] would create therisk of holding the named defendants liable . . . for moreharm than they caused. The . . . theory does not protecttortfeasors from liability exceeding their responsi-bility . , . 114

113. Payton, 386 Mass. at 571, 437 N.E.2d at 188 (citations omitted).114. Id. at 573, 437 N.E.2d at 189. Some readers might recall that this same

problem was presented in a series of cases in the early 1950's, at which time it wasresolved with little fanfare in favor of group responsibility. Those cases dealt withthe problem of apportioning damages when several independent tort-feasors hadinflicted closely successive or contemporaneous injuries on the plaintiff. The clas-sic situation involved multiple polluters. See, e.g., Phillips Petroleum Co. v.Hardee, 189 F. 2d 205 (5th Cir. 1951); Rusch v. Phillips Petroleum Co., 163 Kan.11, 180 P.2d 270 (1947); Landers v. East Tex. Salt Water Disposal Co., 151 Tex.251, 248 S.W.2d 731 (1952). Multiple-vehicle accidents presented another in-stance of the problem of apportioning damages among a number of tort-feasors.See, e.g., Hackworth v. Davis, 87 Idaho 98, 390 P.2d 422 (1964); Maddux v. Don-aldson, 362 Mich. 425, 108 N.W.2d 33 (1961).

In all of these cases, the separate injuries were often so similar and intermin-gled that while negligence and causation could easily be proven against each de-fendant, the plaintiff often had great difficulty proving which part of his damagewas caused by each defendant. The traditional rules of causation required suchproof; otherwise, one individual might be held responsible for harm done by an-other-a violation of the IR principle. See Maddux, 362 Mich. at 438, 108 N.W.2dat 39 (Carr, J., dissenting) (objecting to the majority's departure from the tradi-tional rules); Sun Oil Co. v. Robicheaux, 23 S.W.2d 713, 715 (Tex. Ct. App.1930); RESTATEMENT OF TORTS § 881 (1934). However, beginning with Landers in1952, several state and federal courts concluded that the traditional apportion-ment rule was manifestly unjust and instead adopted a rule that held defendantsjointly and severally liable. Some courts treated this as a rebuttable presumptionthat each individual defendant could rebut with proof of apportionment. See, e.g.,Michie v. Great Lakes Steel Div., Nat'l Steel Corp., 495 F.2d 213, 215-18 (6thCir.), cert. denied, 419 U.S. 997 (1974); Maddux, 362 Mich. at 436, 108 N.W.2d at38; Landers, 151 Tex. at 256-57, 248 S.W.2d at 734.

This new rule was a fairly clear example of group responsibility in an areawhere the law had previously held fast to the IR principle. Indeed, Rosenbergcites Michie as an example of group responsibility. See Rosenberg, supra note 3, at884 n. 139. However, the rule gradually spread to a considerable number ofjuris-dictions. See W. PROSSER & P. KEETON, supra note 1, § 52, at 350-51 & n.52; Note,Torts-Successive Automobile Collisions-Joint and Several LiabilitT, 44 N.C.L. REV. 249,251 (1965); Note, Torts-Liability--Independent Tortfeasors Jointly and Severally Liablefor Separate Acts of Negligence Where Harm is Indivisible: Holtz v. Holder (Ariz. 1966), 9

1508

1986] BETWEEN TWO WORLDS 1509

Other courts have been quick to recognize that the vari-ous special causation rules not only depart from the IR prin-ciple, but actually impose group responsibility. Forexample, the dissent in a Florida appellate court decisionnoted that "the theory espoused in Sindell has the effect ofmaking every manufacturer an insurer not only of its ownproduct, but of all generically similar products made byothers."' 15 Perhaps the clearest statement of this recogni-tion was made by a New Jersey court in Namm v. Charles E.Frosst & Co., 116 a DES case brought on an alternative liabilitytheory:

The application of the principle of alternative liability toany one or all of the 44 defendants herein would imposeliability without fault upon any one who manufactured aproduct manufactured by others as well. It would resultin the taking of the property of all the named defendantsin order to pay for harm which may have been caused by

ARZ. L. REV. 129, 131-32 (1967). Furthermore, this shift from individual togroup responsibility evoked no flood of criticism or controversy among commen-

tators. See, e.g., Note, Recent Developments in Joint and Several Liability, 14 BAYLOR L.REV. 421, 425, 435-37 (1962); Note, Torts-Independent Tort Feasors-Joint and Sev-

eral Liability, 31 N.C.L. REV. 237, 241 (1953). On the contrary, for years prior tothe first cases commentators had loudly insisted upon such a change. See Landers,

151 Tex. at 254-55, 248 S.W.2d at 733;Jackson,Joint Torts and Several Liability, 17TEX. L. REV. 399 (1939); Prosser, Several Liability, 25 CALIF. L. REV. 413 (1937).

If courts and commentators all accepted the shift to group responsibility inthe context of apportionment of damages, why then should its recurrence in thecontext of the causation issue engender so much resistance? The answer is that

some observers regard the modern rule of apportionment of damages as simply adifferent version of the traditional rule that when multiple independent tort-feasors inflict an indivisible injury, they are held jointly and severally liable be-cause each is responsible for a harm indivisible from any other, hence for theentire harm. See, e.g., Maddux, 362 Mich. at 450, 108 N.W.2d at 45-46 (Black, J.,concurring). Even if the indivisible harm rationale is difficult to accept in somecases, the modern apportionment rule is still distinguishable from the group caus-ation decisions. It is one thing to say that when several negligent defendantscause the plaintiff harm, we will sometimes allow the plaintiff to recover, even if hecannot prove the precise share or portion of harm done by each defendant.Whichever defendant or defendants pay, we at least know that those who pay actu-ally injured the plaintiff. It is significantly different to say that when only one ofseveral defendants injured the plaintiff, each can be held liable for some or all ofthe damages. The point here is precisely that we do not know that the defendantwho pays actually injured the plaintiffat all. In other words, although the IR prin-

ciple technically may be violated in both cases, the violation is much more openand obvious in the group causation situation.

115. Copeland v. Celotex Corp., 447 So. 2d 908, 922 (Fla. Dist. Ct. App. 1984)

(Nesbitt,J., concurring in part and dissenting in part), rev"d in part and modified, 471So. 2d 533 (Fla. 1985).

116. 178 N.J. Super. 19, 427 A.2d 1121 (App. Div. 1981).

UCLA LA W REVIEW [Vol. 33:1473

only one of the defendants, or even by one who is not aparty to the lawsuit, who is unknown to the defendants,over whom they have no control or even any meaningfulcontact. '1

7

The court's reaction was also clear: "We reject the theory• . .on principle. ' 18 Of course, the principle on which it isrejected is the IR principle. Other courts, in similar kinds oftoxic exposure cases, have rejected the special causationrules as radical departures from traditional concepts of cau-sation and tort liability." t9 These decisions clearly rest onthe courts' commitment to preserving the IR principle.12o

117. Id. at 33, 427 A.2d at 1128; see also Abel v. Eli Lilly & Co., 94 Mich. App.59, 90-92, 289 N.W.2d 20, 33 (1979) (Moore, J., dissenting), modified, 418 Mich.311, 343 N.W.2d 164, cert. denied, 469 U.S. 833 (1984).

118. Namm, 178 N.J. Super. at 34, 427 A.2d at 1128.119. See, e.g., Thompson v. Johns-Manville Sales Corp., 714 F.2d 581 (5th Cir.

1983) (rejecting alternative, enterprise, and market share liability in a case involv-ing injuries caused by exposure to asbestos), cert. denied, 465 U.S. 1102 (1984);Hannon v. Waterman S.S. Corp., 567 F. Supp. 90 (E.D. La. 1983) (rejecting alter-native liability, concert of action, enterprise liability, and market share liability inan asbestos case); Tidier v. Eli Lilly & Co., 95 F.R.D. 332 (D.D.C. 1982) (rejectingmarket share liability in a DES case); Morton v. Abbott Laboratories, 538 F. Supp.593 (D. Fla. 1982) (rejecting alternative liability, concert of action, enterprise lia-bility, and market share liability in a DES case); Mizell v. Eli Lilly & Co., 526 F.Supp. 589, 596 (D.S.C. 1981) (rejecting market share liability in a DES case on thegrounds that under this rule "liability is placed on defendants bearing no respon-sibility for the defective product"); Zafft v. Eli Lilly & Co., 676 S.W.2d 241 (Mo.1984) (rejecting alternative liability, concert of action, enterprise liability, andmarket share liability theories in a DES case); see also Downey & Gulley, supra note89, at 191 & n.114, and cases cited therein.

Even in California, where Sindell was decided, the courts have refused to ex-tend market share liability. In a case in which plaintiff was injured by a defectivelyproduced batch of polio vaccine and could not prove which of several producershad produced the defective batch, a state appellate court held that it would notuse market share liability to hold the entire producer group responsible. Sheffieldv. Eli Lilly & Co., 144 Cal. App. 3d 583, 597, 192 Cal. Rptr. 870, 878 (1983)(indicating that neither Summers nor Sindell can be used to create "a shared liabilityindiscriminately imposed on manufacturers of safe and defective products of thesame [generic] nature"). The loss spreading, deterrence, and compensation ratio-nales of the Sindell decision actually would seem to favor extending the rule tocover a defective unit of specifically unidentifiable origin. Nor would this involveany greater an extension of the group responsibility premise in theory. Somehave even read Sindell as approving imposition of liability on "the careful andcareless producer alike." See Note, supra note 3, at 317. Yet the Sheffield courtspecifically refused to do just this. See 144 Cal. App. 3d at 597, 192 Cal. Rptr. at878. The only explanation of the Sheffield court's refusal to extend Sindell is that todo so would represent too blatant an adoption of group responsibility. The courtsimply gagged on a more obvious version of what the higher court swallowedquite readily in Sindell. Loyalty to the IR principle apparently dies hard. See infranote 294.

120. This is not to say that the decisions offer no other bases for rejecting the

1510

1986] BETWEEN TWO WORLDS 1511

While some commentators have accepted the groupcausation rules on a "necessity" rationale, many others havesupported the courts' general rejection of these rules andhave severely criticized them. The key reason for this resist-ance has again been commitment to the IR principle:"Holding manufacturers liable for accidents for which theyare in no way responsible, merely because they sold thesame product, offends the notion that liability must bearsome connection to [individual] responsibility . ,"121 An-other commentary quotes approvingly from the dissentingopinion in a Michigan DES case that upheld alternative lia-bility: "Michigan law does not support 'collective liability.'* . . Injury alone does not justify imposing liability on onemanufacturer for injuries caused by another manufacturer ofthe same or similar product."1 22 Loyalty to the IR principle

group causation rules. Frequently, courts argue that these rules, by increasing theindividual producer's exposure to liability, will have an inhibiting effect on thedevelopment and marketing of new drugs and other useful but potentially injuri-ous products. See Sindell, 26 Cal. 3d at 619, 607 P.2d at 942, 163 Cal. Rptr. at 146(Richardson, J., dissenting); Payton, 386 Mass. at 571-75, 437 N.E.2d at 189-90.However, this concern about overdeterrence is not clearly a logical basis for re-jecting group causation. On the one hand, greater deterrence is one of the argu-ments often made in favor of the group causation rules, see supra notes 108-10 andaccompanying text. On the other hand, some critics argue that group causation isundesirable because it will shield individual wrongdoers and thus provide too littledeterrence. See Note, supra note 3, at 315-21. But see infra note 254 and accom-panying text for the view that group causation rules do increase deterrence.Therefore, the deterrence issue is much less clear than the IR principle as a basisfor favoring or disfavoring the new group causation rules.

121. Note, supra note 3, at 327.122. Downey & Gulley, supra note 89, at 191 (citing Abel v. Eli Lilly & Co., 94

Mich. App. 59, 90-92, 289 N.W.2d 20, 33 (1979) (MooreJ., dissenting) (buttress-ing this conclusion by observing that "one may not be obligated for the fault of afellow lawyer, or doctor, or engineer, or a drug manufacturer"), modified, 418Mich. 311, 343 N.W.2d 164, cert. denied, 469 U.S. 833 (1984)). Several authorsconclude that the group causation rules in effect institute a nonfault compensationplan for drug injuries by judicial fiat. Recognizing that nonfault compensationplans generally abandon the IR principle, see supra note 89, these authors object tosuch a radical departure from traditional torts principles. In the alternative, theysuggest that only legislative action can legitimately support such a departure, as inother compensation schemes. See Downey & Gulley, supra note 89, at 195-200;Note, Manufacturers' Liability, supra note 106, at 315; Note, Products Liability:Sindell v. Abbott Laboratories: Proportional Unidentifiable Fairness and the OklahomaPerspective, 34 OKLA. L. REV. 843, 861 (1981); Note, Eliminating the Identification Re-quirement, supra note 89, at 627-28; see also Note, Best Remedy, supra note 89; Note,Beyond Enterprise Liability in DES Cases-Sindell, 14 INn. L. REV. 695 (1981) [herein-after cited as Note, Beyond Enterprise Liability]; Comment, Market Share LiabilityAdopted To Overcome Defendant Identification Requirement in DES Litigation, 59 WASH.

U.L.Q 571 (1981) [hereinafter cited as Comment, Market Share Liability Adopted].

1512 UCLA LAWREVIEW [Vol. 33:1473

leads these commentators to resist and reject the group cau-sation rules. 123

However, the legislative response may be equally stymied by commitment to theIR principle. See infra note 134.

123. Some readers may object that this loyalty is misplaced and unrealistic, ar-guing as follows. Given the prevalence of liability insurance coverage, "collectiveliability" or group responsibility has long been the reality of tort law, whatever theformal doctrine says. For many decades, tort liability has coexisted with a broadlyutilized system of accident liability insurance. The insurance system itself oper-ates in large part as a risk-pooling device, grouping together individuals with simi-lar accident-risk profiles and charging each individual ex-ante an amountproportional to his share of the aggregate risk, or expected accident cost, of thegroup as a whole. When an accident occurs and an insured individual is heldliable under traditional tort rules based on the IR principle, the actual cost ofliability nevertheless is shared with other members of his risk pool because theaward is paid out of the funds generated by all the ex-ante contributions. Simi-larly, whenever any other member of the "relevant actuarial category" is heldliabile for inflicting an injury, every insured individual will pay a share of the dam-ages through his premium. In short, whenever a defendant is insured, the insur-ance system, by its very nature, converts individual responsibility into groupresponsibility. For a discussion of this phenomenon, and an argument that it isnot only desirable but also necessary, see G. CALABRESI, supra note 3, at 60-64,90-92, 103-07. For a discussion of the distinction between liability insurance andsocial insurance, which involves societal or collective rather than group responsi-bility, see supra note 106.

A similar analysis can be made from the victim's perspective. For most vic-tims of insured defendants, the insurance settlement process determines theircompensation. In this process, settlement offers are standardized by reference tocategories and types of accidents and injuries. The individual is thus functionallytreated as a member of a group of similarly injured individuals, each of whomreceives an offer that roughly reflects his share of the aggregate sum set aside forthat kind of injury in a given period. See H. Ross, SETTLED OUT OF COURT 114-16,133-35 (1980); Rosenberg, supra note 3, at 909 n.229. Some victims get windfalls,while others receive shortfalls. Therefore, the character of the system, from boththe victim's and defendant's perspective, is much like Delgado's and Rosenberg'sgroup responsibility proposals. See supra text accompanying notes 38-51. Giventhe widespread and accepted utilization of liability insurance and insurance settle-ments, why should group causation rules in tort law be considered controversial,when the system has functionally involved group responsibility for decades, de-spite the individualistic form of the legal rules?

The response to this question illuminates the nature of the current contro-versy over the group causation rules and proposals. The liability insurance sys-tem, though it represents a form of group responsibility, does not present thekind of open and obvious challenge to the IR principle that the group causationrules present. The liability insurance system is a creature of private enterprise,not a rule of public law. It is acceptable for individuals to choose on their own toshare risks; it is altogether different when individuals are forced by the state toshare responsibility for the actions of others. Indeed, when the insurance systemhas taken on a public character, as in laws compelling insurance coverage, or lawsestablishing or proposing administrative compensation schemes functionallyequivalent to state sponsored liability insurance, there has been significant contro-versy and resistance. See, e.g., Blum & Kalven, The Empty Cabinet of Dr. Calabresi:Auto Accidents and General Deterrence, 34 U. Cm. L. REV. 239 (1967); Blum & Kalven,

1986] BETWEEN TWO WORLDS I1 i3

The judicial and scholarly resistance has not been lim-

ited to the group causation rules themselves. It has ex-

tended to the risk contribution and lost value theories, both

of which clothe group responsibility in a more individualisticform. The source of the resistance in this context can also

be found in the courts' concern for the IR principle. For

example, in Jackson v. Johns-Manville Sales Corp., 124 a plaintiff

exposed to asbestos sued for damages representing his in-

creased risk of future cancer, based on epidemiologic stud-

ies. While Robinson's and King's proposals endorse

precisely this kind 'of recovery, the court rejected plaintiff's

claim: "To permit this assumption of causation to control

liability would cause massive dislocation in the rights of the

parties . . . . '[Some] [p]ersons ...will receive a windfall

and, in the aggregate, the defendant will overcompensatethe injured class.' "125 The court was bothered by precisely

the aspects of the risk contribution theory that involve rec-

ognizing responsibility to a group, rather than to particularindividuals. In essence, the court rejected the shift from in-

dividual to group responsibility. 126

Basing responsibility on purely probabilistic evidence

A Stopgap Plan for Compensating Auto Accident Victims, 1968 INS. L.J. 661 (1968);

Schwartz, Professor O'Connell's No-Fault Plan for Products and Services: Have New

Problems Been Substituted for Old?, 70 Nw. U.L. REV. 639 (1976); Atiyah, Book Re-

view, 45 U. CIN. L. REV. 340 (1976) (reviewingJ. O'CONNELL, ENDING INSULT TO

INJURY (1975)).

In short, there is an open and obvious challenge to the IR principle inherent

in the group causation rules that is absent in the private insurance process (and in

the modern rule of apportionment of damages, see supra note 114); and this chal-

lenge has evoked resistance and rejection. As long as the IR principle was left

substantially and formally intact as a limit on what kind of responsibility society

can impose forcibly, there was no real challenge and no need for a response. But

when a change in a major legal rule openly declared that society can in some

cases, perhaps many, forcibly impose group responsibility, then the challenge wasissued in no uncertain terms. The response likewise has been loud and clear. Seenotes 111-34 and accompanying text; see also supra note 97.

124. 727 F.2d 506 (5th Cir. 1984).

125. Id. at 520 (quoting Wilson v. Johns-Manville Sales Corp., 684 F.2d 111,

120 n.45 (D.C. Cir. 1982)).

126. In a similar case involving an action for increased cancer risk from pesti-

cide exposure, a California court concluded on the basis of scientific and medical

literature that "as a matter of law it cannot be determined with any degree of

medical probability whether a particular plaintiff... will contract cancer in the fu-ture as a result of his exposure .... Arnett v. Dow Chem. Co., 6 Chem. &

Radiation Waste Litig. Rep. (BNA) 383, 390 (Cal. Super. Ct. 1983). The court

was uncomfortable allowing a plaintiff who was unable to identify himself individu-

ally as a victim of the defendant's actions to recover simply by proving his mem-

UCLA LAW REVIEW

involves a form of group responsibility. This undoubtedlyexplains the widespread judicial suspicion of probabilistic orstatistical evidence as sufficient to meet the causation re-quirement in a torts action. 2 7 For example, the UtahSupreme Court, in the widely cited case of Garner v. HeclaMining Co., 128 rejected a uranium miner's claim for damagesfrom radiation-caused cancer, which he based on the strongstatistical correlation between exposure and incidence of thedisease:

While it seems logical that the unusually high incidenceof lung cancer in uranium miners would indicate in thesame ratio the higher probability than otherwise thatsuch was the cause of the disease, it nevertheless fallsshort of compelling a finding that such was the cause inany individual case. 129

The court was unwilling to determine the defendant's liabil-ity on any basis other than responsibility to an individual vic-tim, which simply could not be proven. Such statisticalevidence does prove responsibility for an aggregate injury toa group of uncertain membership, but courts generally havebeen very reluctant to hold defendants liable on this basis,for this would be a form of group and not individualresponsibility.

In perhaps the best-known critique of probabilistic evi-dence, Professor Tribe ultimately bases his manifold objec-tions to the use of statistical evidence upon "our society'straditional affirmation of the 'presumption of innocence,' "and its underlying values of "affirm[ing] the dignity of theaccused," "displaying respect for his rights as a person,"and "refusing to sacrifice him to the interests of others."' 130

Statistical evidence leads us to pay too little attention to theindividual because it shifts our attention from his rights and

bership in a group, some of whose members very likely suffered or would sufferinjury from the defendant's actions.

127. See Dore, supra note 65, at 430 & n. 11; Robinson, supra note 12, at 764-65;Rosenberg, supra note 3, at 857; Wright, supra note 2, at 1821-26; Note, JudicialAttitudes Towards Legal and Scientific Proof of Cancer Causation, 3 COLUM. J. ENVTL. L.344, 355-70 (1977); Note, Tort Actions for Cancer, supra note 38, at 847-48 & n.32,851 & n.51, 854 & n.65, 858 & n.78. But see Dore, supra note 65, at 434 &nn.43-46 for cases holding otherwise.

128. 19 Utah 2d 367, 431 P.2d 194 (1967).129. Id. at 370, 431 P.2d at 796 (emphasis added).130. Tribe, Trial By Mathematics: Precision and Ritual in the Legal Process, 84 HARV.

L. REV. 1329, 1370, 1374 (1971).

1514 [Vol. 33:1473

1986] BETWEEN TWO WORLDS 1515

responsibilities to those of some statistically determinedclass. Others have made the point in closer connection totort law.1 31 Criticizing Calabresi's work,1 32 one commenta-tor rejects the notion that "causal linkage," essentially an-other term for statistical correlation, is sufficient to establishcausation: "It [causal linkage] obtains only between classes ofconduct and harm .... The causal relation that is pertinentin a system of corrective justice must exist between the [par-ticular] defendant's conduct and the [particular] plaintiff'sharm."' 33 For these critics, reliance on statistical correlationto prove causation is unacceptable because it would be ashift from individual to group responsibility. 34

131. See, e.g., Borgo, Causal Paradigms in Tort Law, 8J. LEGAL STUD. 419 (1979).

132. Calabresi, supra note 78.133. Borgo, supra note 131, at 424-25.134. The rejection of group causation theories and proposals occurs just as

consistently in the legislative arena as in the judicial arena. As noted earlier, seesupra notes 89, 122, many commentators have argued in favor of legislative actionto establish nonfault compensation schemes that would effectively involve groupresponsibility. Despite the substantial number of legislative proposals actually in-troduced, lawmakers have refused to enact them. See Downey & Gulley, supra note

89, at 196-200; Note, Inapplicability of Traditional Tort Analysis, supra note 38, at589-91. As Rosenberg has observed, "[legislatures continue to rely on the tortsystem .. .[and] have declined the invitation to formulate a comprehensive ad-ministrative solution-in part because ... such a solution seems politically infeasi-

ble .... Congress has consistently rejected comprehensive plans." Rosenberg,

supra note 3, at 926 & n.283. I suggest that the political barriers to the compensa-

tion schemes reflect a concern for the IR principle among both legislators and

their constituents. There probably is little support for proposals that would taxindividuals to pay for injuries that they did not personally cause, and offer themless compensation than they might recover in the traditional tort system.

While there has been little if any legislative action on comprehensive com-

pensation schemes, a growing number of states have adopted legislation involvingtort reforms of other kinds. One increasingly common reform measure that might

seem related to the subject of this Article is limitation or elimination ofjoint andseveral liability in cases of multiple tortfeasors, especially for noneconomic dam-

ages such as pain and suffering. Commonly, the state statutes being adopted limit

each defendant's liability for noneconomic damages to a pro rata or equitableshare of the total damages figure. See, e.g., N.Y. CIv. PRAC. LAw §§ 1600-1603

(McKinney 1986). For a summary of recent legislation in this area, see INS. INFO.INST., The Civil Justice System, in DATA BASE REP. 1-2, 9-11, 13-15 (R. Gastel ed.

Oct. 1986). These measures show a tendency on the part of legislatures to be

concerned, like the Sindell court, about the burden ofjoint and several liability onmultiple tortfeasors and the need to place some limits on liability. See supra text

accompanying notes 33-35. However, these measures do nothing whatsoever to

change the basis of liability from individual to group responsibility. Legislatureshave been as resistant as courts to making such a fundamental change.

UCLA LA W RE VIEW

2. Resistance to Group Causation

Perhaps the most interesting examples of judicial andscholarly resistance to rules and proposals involving groupresponsibility are decisions and articles in which the resist-ance is covert rather than overt. In these cases, the authorsadopt a rule or proposal involving a form of group responsi-bility, but then retreat from their position by reintroducingelements of individual responsibility, almost as though theyfear they have gone too far. Such ambivalence toward groupresponsibility, even among some of its most forthright pro-ponents, speaks volumes about the grip of the IR principleon American jurisprudence.

For example, in Allen v. United States, 135 the case involv-ing injuries allegedly caused by nuclear radiation, the UtahFederal District Court appears to adopt the rule that proba-bilistic evidence is sufficient to establish individual causa-tion. 136 Especially where the result is full recovery, as it is inAllen, such a rule would involve a form of group and not in-dividual responsibility as discussed earlier. However, havingexplicitly rested its holding on several theories of group re-sponsibility, including Delgado's indeterminate plaintiff the-ory and Calabresi's notion of causal linkage, the Allen courtthen retreated from its initial position.13 7 The court pro-ceeded as trier of fact to consider individually, with refer-ence to each of the twenty-four plaintiffs, "proof ofadditional factual connections which indicate that exposureto fallout materially augmented the plaintiffs' risk of injury,which took effect in actual somatic injury . ,,138 Based onthese individual determinations, the court awarded damagesto only ten of the twenty-four plaintiffs. 139 This is hardly theresult that one would expect under a group responsibilityapproach to the situation. Allen is a far cry from adoptingrisk contribution liability, or any other form of group re-sponsibility, despite the clear language in the opinion sug-

135. 588 F. Supp. 247 (D. Utah 1984).136. See supra text accompanying notes 76-82.137. See 588 F. Supp. at 407, 414-16.138. Id. at 430.139. See id. at 427-47. Many of the alleged radiation victims were deceased,

and their claims were actually litigated by spouses, children, and parents. In theopinion, the court referred to the alleged victims themselves as "plaintiffs" and tothe survivors as "claimants." See id. at 427-28, 446-47. Recovery was ultimatelyallowed to claimants representing only 10 of the 24 plaintiffs.

1516 [Vol. 33:1473

BETWEEN TWO WORLDS

gesting adoption of this rule. 140 The court was simplyunwilling to ignore the IR principle's emphasis upon the in-dividual link between injurer and victim.

Ironically, the very commentators that appear to be thestrongest supporters of group responsibility actually demon-strate ambivalence or covert resistance to the principle. Ro-senberg, for instance, seems to be unequivocal in hissupport of an explicitly "aggregative" and class-oriented ap-proach to causation in torts. 14 1 Nevertheless, within thesame breath, he devotes considerable effort to arguing that"individual entitlements" would be preserved better by pro-portional liability than by the traditional causation rules. 142

Furthermore, after stressing the "collective" nature of thevictims' interest in a mass exposure case, and the necessityfor class treatment, 43 he immediately switches ground andargues that "the class action preserves the most individual-ized component of the trial-the phase devoted to assessingthe [individual] victim's loss . . . . Class actions ... thusmake possible . . . individualized assessment of loss."' 144

However, this argument is totally disingenuous, as Rosen-berg himself later confirms in his discussion of scheduleddamages. Such damages, he concedes, "award compensa-tion to claimants not on the basis of their personal character-istics, but rather on the basis of characteristics of a class ofwhich the individual was a member."1 45 This hardly involvesindividualized assessment of loss!

Actually then, Rosenberg is not unequivocal in his sup-port for group responsibility. Perhaps Rosenberg genuinelyretains a commitment to the IR principle. This is suggestedby his obvious acceptance of the corrective justice theory oftort law as one clear standard by which his proposal must bejustified. 146 On the other hand, perhaps he anticipates a

140. Allen actually represents a relatively conservative use of what Rosenberg

calls the "strong version" of the preponderance rule, which requires some partic-

ularized additions to statistical correlations before allowing recovery on that basis.

See Rosenberg, supra note 3, at 857, 869.

141. See Rosenberg, supra note 3, at 855, 859-60, 884-85, 905-08; see also supratext accompanying notes 38-50.

142. See Rosenberg, supra note 3, at 877-81.

143. See id. at 907.

144. Id. at 909-10.145. Id. at 917.146. See id. at 877-87.

151719861

1518 UCLA LAWREVIEW [Vol. 33:1473

negative reaction to his proposal because of "apprehensionabout the collectivist dangers" of group responsibility.' 4 7

Consequently, he may be trying to forestall this reaction byasserting that his scheme somehow supports the IR principleafter all. Either way, the IR principle remains strong enoughto influence even this foremost proponent of a group re-sponsibility approach in tort law and to lead him into incon-sistencies and contradictions."4 8

147. Id. at 909.148. The same ambivalence may underlie the efforts of other commentators,

like Robinson and King, to clothe what are essentially group responsibility pro-posals in forms that appear to reflect individual responsibility. See supra text ac-companying notes 52-65.

The asbestos cases present a third example of implied ambivalence towardgroup responsibility. In these cases, plaintiffs sued for injuries caused by expo-sure to asbestos over a period of time and were unable to identify which pro-ducer's asbestos caused their injury. In the majority of jurisdictions that havefaced the issue, including California, the courts have rejected market share liabil-ity. See Thompson v. Johns-Manville Sales Corp., 714 F.2d 581 (5th Cir. 1983),cert. denied, 465 U.S. 1102 (1984); Hannon v. Waterman S.S. Corp., 567 F. Supp.90 (E.D. La. 1983); In re Related Asbestos Cases, 543 F. Supp. 1152 (N.D. Cal.1982); Starling v. Seaboard Coast Line R.R., 533 F. Supp. 183 (S.D. Ga. 1982);Prelick v. Johns-Manville Sales Corp., 531 F. Supp. 96 (W.D. Pa. 1982); see alsoCopeland v. Celotex Corp., 447 So. 2d 908, 917-22 (Fla. Dist. Ct. App. 1984)(Nesbitt, J., concurring in part and dissenting in part) (agreeing with the major-ity's conclusion that the plaintiff's complaint alleged a sufficient cause of action,but rejecting the majority's reliance upon Sindell). Significantly, the majority'sopinion in Copeland was modified by the Florida Supreme Court. See Celotex v.Copeland, 471 So. 2d 533, 537 (Fla. 1985) ("Market share liability is an inappropri-ate vehicle with which to apportion liability for the asbestos-related injury in thiscause."). In contrast to the substantial number of asbestos cases rejecting marketshare liability, few courts have upheld the theory. See Hardy v. Johns-ManvilleSales Corp., 509 F. Supp. 1353 (E.D. Tex. 1981), rev'd, 681 F.2d 834 (5th Cir.1982).

In part, the courts' justification for rejecting the Sindell doctrine stems fromtheir application of a special rule that has evolved in asbestos cases. When it isproven that a defendant's product caused some portion of the plaintiff's injury,the identified producer can be held fully liable for plaintiff's entire harm, becausethe injury inflicted by one manufacturer's product is considered "indivisible"from any injury inflicted by the products of others. See, e.g., Borel v. FibreboardPaper Prods. Corp., 493 F.2d 1076, 1094-96 (5th Cir. 1973), cert. denied, 419 U.S.869 (1974). Because the courts have held that there is at least one identifiableinjurer in most asbestos cases, Sindell and other special causation rules are unnec-essary and inapplicable. See Starling, 533 F. Supp. at 191; Prelick, 531 F. Supp. at98; see also Copeland, 447 So. 2d at 917 (Nesbitt, J., concurring in part and dissent-ing in part). Ironically, this position can result in an even stronger form of groupresponsibility than Sindell: a sole identified producer can be held responsible forinjuries inflicted by all other producers. Like the single defendant in Collins v. EliLilly & Co., 116 Wis. 2d 166, 342 N.W.2d 37, cert. denied, 469 U.S. 826 (1984), the

BETWEEN TWO WORLDS

C. The Roots of the IR Principle: The Political Theory ofLiberal Individualism

The previous two sections have shown that the IR prin-ciple underlies a widespread resistance to the causation rulesand theories discussed in Part I that involve a shift to groupresponsibility. As indicated in the foregoing sections, anever-increasing number of cases involve problems of inde-terminate causation. Nevertheless, most courts refuse to re-spond to these problems in any way that could compromisethe IR principle. The obvious question is, why has the gripof the IR principle remained so strong? Why is it able tosupport this resistance despite major pressures for change?One answer, as suggested briefly in the Introduction, is thatthe IR principle underlies not only the traditional causationrequirement, but also the theories of utilitarianism and cor-rective justice most commonly offered to explain the socialfunction of tort law as a whole. 149 To challenge the principleon one level by abandoning the traditional causation re-quirement therefore challenges the larger theoretical struc-tures as well and evokes great resistance.

There are two problems with this answer. First, the util-itarian, or deterrent, theory of torts is generally argued ineconomic terms which themselves involve analysis of con-

duct by classes or groups of actors, not individuals. 150 Inother words, the utilitarian analysis itself implies that the de-terrent function of torts can be served very effectively by im-posing responsibility on groups of actors.isi If so, thendespite superficial impressions to the contrary, there is sim-

asbestos producer can be held responsible for the injuries inflicted by every other

member of his group. See supra notes 66-75 and accompanying text.

Unlike Collins, however, the asbestos cases that reject Sindell do not suggestthat they are purposely extending group responsibility in order to ensure fullplaintiff recovery. On the contrary, their rejection of market share liability seems

tied to the simple conviction that individual identification, whenever it can be

made, is preferable to group identification. This does not logically support the IR

principle, and actually may involve a greater deviation from that principle than

was involved in Sindell. But this position has the appearance of preserving individ-

ual responsibility by avoiding open identification by group membership. In the

asbestos cases, the desire to maintain the appearance of individual responsibilityhas had the paradoxical effect of supporting a rule that may actually impose a

more pronounced form of group responsibility.

149. See supra notes 3-5 and accompanying text; see also infra notes 170, 186.

150. See supra note 107.

151. See, e.g., Calabresi & Hirschoff, supra note 96, at 1067-70, 1075.

15191986]

UCLA LAW REVIEW

ply no conflict, on a functional level, between the utilitariantheory of torts and open adoption of group responsibility.Actually, some economic and utilitarian theorists have longadvocated adopting group responsibility as a key element intheir proposals for overhauling the torts system as awhole. 52 In addition, some courts and commentators rec-ognize increased deterrence as an argument in favor of thenew group causation rules.' 53 Thus, group responsibilityhardly undermines the utilitarian theory of torts.

On the other hand, it is quite true that any principleother than individual responsibility is incompatible with thehighly individualistic focus of the corrective justice theory onthe connection between a particular victim and a particularinjurer. 54 According to this theory, the function of tort lawis to ensure that "as a matter of individual justice betweenthe plaintiff and the defendant, the defendant who hascaused an injury to the plaintiff.., must compensate him...whether or not [this] . . .will further some collective socialgoal." ' 55 Any rule based upon group responsibility wouldobscure or destroy the individual injurer-victim connection;hence, it would obviate the whole function of tort law ac-cording to the corrective justice theory. Because any chal-lenge to the IR principle challenges the corrective justicetheory of torts, the commitment to this theory may be thereal source of the resistance to group causation. However,this only broadens rather than answers the original question:Why is the grip of both the IR principle, and the correctivejustice theory of torts that it supports, so strong?

152. See, e.g., G. CALABRESI, supra note 3, at 246-50, 255-59, 282-83.153. See supra text accompanying notes 106-10; Sindell v. Abbott Laboratories,

26 Cal. 3d 588, 610-13, 607 P.2d 924, 936-37, 163 Cal. Rptr. 132, 144-45, cert.denied, 449 U.S. 912 (1980); Collins v. Eli Lilly & Co., 116 Wis. 2d 166, 192-93,342 N.W.2d 37, 49-50, cert. denied, 469 U.S. 826 (1984).

154. See Borgo, supra note 131, at 424-25, 444-45; Englard, supra note 3, at 27& n.2; Rosenberg, supra note 3, at 877; Wright, supra note 3, at 435-36, 455-56;Note, Inapplicability of Traditional Tort Analysis, supra note 38, at 605-06. Rosen-berg's restatement of the corrective justice basis of the causation requirement ob-scures the importance of the individual victim-injurer connection. When he says"the causal connection is a necessary ... pre-condition to liability," Rosenberg, supranote 12, at 877 n.108 (emphasis added), his choice of terminology deemphasizesthe concern of corrective justice for rectification of the situation between the par-ticular victim and injurer. See id. at 855. Rosenberg's notion of "causal connec-tion," see id. at 855 & n.27, is similar to Calabresi's "causal linkage" and can applyto groups of injurers and victims. See supra text accompanying notes 131-33.

155. Wright, supra note 3, at 435.

1520 [Vol. 33:1473

BETWEEN TWO WORLDS

This question cannot be answered from within the con-fines of legal doctrine and theory per se. An answer must besought in the larger context of political structure and theory,of which law is a part. 15 6 In his historical study of the tradi-tional causation requirement, Professor Zwier notes that"It]he burden of proving cause in fact [arose] from . ..asystem of laws that placed primary emphasis on the rights ofthe individual." 157 That system of laws, including the IRprinciple and the corrective justice view of tort law, in turnderived from "an underlying conception of a society of indi-viduals in which each individual possesses equal liberties,autonomy, and freedom."' 1 8 In political theory, this con-ception is generally referred to as the theory of liberalism, orliberal individualism. 159

The focus of liberal theory is the individual and hisfulfillment:

Society is composed of individuals. It has no reality in-dependent of its members. What appear to be the ac-complishments of the whole are in fact the sum of worksof individuals. Groups exist; but they reflect a conver-gence of separate interests. And the group is not asource of value in its own right. For the good is relativeto the wants of distinct persons. The immediate measureof conduct lies within the individual rather than in thegroup to which he belongs. 160

Political philosophers and historians locate the origins ofthis theory in the 17th- and 18th-century attack on the aris-tocratic and status-based society of feudal and postfeudalEurope.' 6 1 In that earlier order, "[s]ocial arrangementswere . . .based on collective units like the family or the

156. See K. LLEWELLYN, THE BRAMBLE BUSH 107-08 (1960); Kairys, Introduction,in THE POLITICS OF LAW, supra note 2, at 1, 4-5; Note, 'Round and 'Round the Bram-ble Bush: From Legal Realism to Critical Legal Scholarship, 95 HARV. L. REV. 1669,1671-73, 1677-79.

157. Zwier, supra note 2, at 784.158. Note, Inapplicability of Traditional Tort Analysis, supra note 38, at 605 n. 134.159. See M. SANDEL, supra note 15, at 1; Murphy, supra note 16, at 125-26. For

further clarification of the meaning of the terms "liberalism" and "social welfar-ism" as used herein, see supra note 16.

160. Murphy, supra note 16, at 126; see also J. RAWLS, A THEORY OF JUSTICE264-65 (1971).

161. See MacDuff, Causation Theory and Uncertainty, 9 VICT. U. WELLINGTON L.REV. 87, 88 (1978); see also Haney, Criminal Justice and the Nineteenth-Century Para-digm, 6 LAW & HUM. BEHAV. 191, 193-94 (1982); Murphy, supra note 16, at 125,153-54; Zwier, supra note 2, at 791.

1986] 1521

UCLA LAW REVIEW

guild," and "[c]ollective doctrines of legal culpability ...flourished ..... ,162 However, the abuses of the privilegedclasses provoked both philosphical and political revolt, ulti-mately leading to the creation of the modern democraticstate. During that process, the intermediate collective unitsor groups that had originally provoked the demand forchange were largely swept away, their legitimacy irrevocablycompromised by the abuses they engendered.

At the time, political thinkers saw the result as "theachievement by the self of its proper autonomy. The selfhad been liberated from all those outmoded forms of socialorganization .. . . When the distinctively modern self wasinvented . . . [w]hat was then invented was the individual.... "163 The resulting political universe was a dualist one.In both political theory and law, the only cognizable entitieswere the individual and the state, and all legal relations weredefined accordingly. 164 Therefore, infliction of injury couldonly be viewed as an individual injustice. Such injustice de-manded correction, but correction could only occur throughreparation of the individual victim by his individual injurer.Doctrines of collective responsibility in law and elsewhere"were replaced by ones locating primary responsibility in in-dividuals."'' 65 At the same time, the state itself derived itslegitimacy from the individuals who were its constituentsand who alone were sovereign. Therefore state action hadto be precise in matching responsibility to actor; otherwise,the rights of an "innocent" individual might be violated.Thus, apart from the theory of corrective justice, the IRprinciple was justified as a necessary "protection of the indi-vidual from intrusion by the state."' 166

In sum, the IR principle and the corrective justice viewof torts are part and parcel of the liberal theory, according towhich both law and political structure must reflect the ulti-

162. Haney, supra note 161, at 194; see alsoJ. FEINBERG, supra note 11, at 235.163. A. MACINTYRE, AFTER VIRTUE: A STUDY IN MORAL THEORY 58-59 (1981)

(emphasis in original); see also Presser, Some Realism About Orphism, or The CriticalLegal Studies Movement and the New Great Chain of Being. An English Legal Academic'sGuide to the Current State of American Law, 79 Nw. U.L. REV. 869, 878 (1984) (citingM. HOROWITZ, THE TRANSFORMATION OF AMERICAN LAW 1780-1860 (1977)).

164. See Macneil, Bureaucracy, Liberalism, and Community-American Style, 79 Nw.U.L. REV. 900, 917-18 n.65 (1984).

165. Haney, supra note 161, at 194; see also MacDuff, supra note 161, at 88.166. Zwier, supra note 2, at 816.

1522 [Vol. 33:1473

BETWEEN TWO WORLDS

mate value of the individual person as the fundamental unitand basis of society. Society's sole function, on the otherhand, is to facilitate the individual's self-fulfillment.1 67

Rawls states the underlying premise of this theory:The essential idea is that we want to account for the so-cial values ... by a conception of justice that in its theo-retical basis is individualistic .... [W]e do not want torely on an undefined concept of community, or to sup-pose that society is an organic whole with a life of its owndistinct from and superior to that of all its members

168

From this view is derived the fundamental principle of lib-eral, rights-based jurisprudence that we may not violate indi-viduals for the greater social good: "[T]here is no social entitywith a good [of its own].... There are only individual peo-ple . . . with their own individual lives. Using one of thesepeople for the benefit of others ... does not sufficiently re-spect and take account of the fact that he is a separate per-son."' 169 The mirror image of this individual rights principleis the individual responsibility principle. 170

Today, the liberal theory remains the dominant concep-tual framework of modern American political thought. Itsprimary rival has been social welfare theory, which in effectembraces the opposite pole of the dualist universe of mod-ern political theory. Social welfarism maintains that thegood of society in the aggregate is the supreme value, andthe goal of law and social policy should be to maximize col-lective social welfare.' 7' There has been a persistent tensionand accomodation between these rival frameworks in Ameri-can political theory, law, and public policy. Nevertheless, ithas never been considered acceptable that the welfaristframework should dominate, much less negate, the liberalvision of individual sovereignty. This would legitimize a col-lectivist ethos that is basically repugnant to our political tra-dition. The common law in particular has evolved as the

167. See Murphy, supra note 16, at 125, 153.168. J. RAWLS, supra note 160, at 264-65.169. R. NOZIcK, ANARCHY, STATE, AND UTOPIA 32-33 (1974).

170. According to one commentator, all the modern tort theorists, from boththe utilitarian and corrective justice schools, ultimately share the individualisticassumptions of liberalism. See Englard, supra note 3, at 68-69.

171. See M. SANDEL, supra note 15, at 3-7; Wright, supra note 3, at 436-37. Forfurther clarification of the meaning of the terms "liberalism" and "social welfar-ism" as used herein, see supra note 16.

15231986]

UCLA LA W REVIEW

heartland of the liberal theory. 72 The traditional causationrequirement and the IR principle that it embodies are themanifestation in tort law of the deeply rooted political the-ory of liberalism.

To answer the initial question raised in this section,courts and commentators have resisted group causationrules and theories not because of the grip of the IR principleor the corrective justice theory, but rather because of thegrip of the underlying theory of liberalism. This theory con-sititutes the dominant conceptual view of our political andsocial structure as a whole. Whether or not they articulate itin this manner, the courts and commentators who resist thenew rules and theories at least intuitively perceive that a shiftto group responsibility would ultimately undermine the lib-eral theory as a political world view.

It is important, before going further, to state moreclearly how and why group responsibility offends the liberaltheory. This will help explain why courts committed to theliberal theory feel compelled to reject group responsibility,even if it is strongly supported by arguments of increaseddeterrence of injury and better compensation of victims.

D. The Objections to Group Responsibility: Submergence of theIndividual and Collectivism

Group responsibility offends the liberal theory in twomajor respects.173 First, by regarding the individual victimor injurer as essentially a member of some group or associa-tion, group responsibility rejects the most fundamentalpremise of liberalism-that each individual is a separate andsovereign being whose existence and fulfillment, to the ex-tent consistent with the same for others, is the whole raison d'tre of society. 174 Ignoring the unique identity and value of

.

172. As gradual democratization of legislatures posed new, majoritarianthreats to individual rights, the courts emerged as the most important vehicle forcarrying out the liberal theory of society. See Murphy, supra note 16, at 128-30,154. Thus, the political philosophy of liberalism found its clearest expression injudge-made rules of law and legal doctrine. See M. SANDEL, supra note 15; Wright,supra note 3.

173. This section was inspired by a conversation with my colleague, ProfessorJohn D. Gregory.

174. See R. DWORKIN, TAKING RIGHTS SERIOUSLY 198-99, 271 (1977); R.NozICK, supra note 169, at 30-33;J. RAWLS, supra note 160, at 3-4, 264-65; see alsoM. SANDEL, supra note 15, at 47-59; Murphy, supra note 16, at 125-26, 153.

1524 [Vol. 33:1473

BETWEEN TWO WORLDS

the individual, group responsibility submerges the individ-ual in a group or class of "similar" parties and penalizes orcompensates him not as an individual, but as a member of aclass. Recalling Tribe's objections, such treatment fails toprovide an "affirmation of respect," to "affirm the dignity,"and to "respect [the] rights as a person" of the individual. 175

Group responsibility fails in all these respects precisely be-cause it fails to recognize and address the individual as such.Advocates of group responsibility argue that it is all right todisregard the individual because the net result will be moreeffective deterrence and greater compensation.1 76 Yet thisclearly sacrifices the individual to serve the interests ofothers-the antithesis of the most fundamental premise ofliberalism. 177 Moreover, if injury law can disregard individu-als and submerge them in groups, what is to prevent thesame from occuring in other areas of law, politics, or sociallife? Group responsibility in injury causation, however paro-chial a concern, is thus a thread from which the entire fabricof liberalism, and its concern for the liberation of the indi-vidual, can be too easily unraveled. 178

The second objection of liberalism to group responsi-bility is related but distinct. If groups are recognized forpurposes of assigning responsibility, the same can be truefor assigning rights. Indeed, group causation theories in-volve both these notions. A concept of group rights, how-ever, opens the door to abuses by groups of greater wealth,influence, or numbers, who are able to use their grouppower to establish superior "rights" to the detriment ofother groups and vulnerable individuals. 79 At the extreme,

175. Tribe, supra note 130, at 1370, 1374. In effect, group responsibility saysthat society will no longer regard the individual as the fundamental unit of legalrelations. Instead, society will regard groups as fundamental units, and the fate ofindividuals within those groups will no longer be a direct concern.

176. See supra text acccompanying notes 106-10; see also Sindell v. Abbott Labo-ratories, 26 Cal. 3d 588, 611-14, 607 P.2d 924, 936-37, 163 Cal. Rptr. 132,144-45, cert. denied, 449 U.S. 912 (1980); Collins v. Eli Lilly & Co., 116 Wis. 2d166, 190-94, 342 N.W.2d 37, 49-50, cert. denied, 469 U.S. 826 (1984).

177. See R. DWORKIN, supra note 174, at xi, 92, 146-47, 269-70; R. NozICK,supra note 169, at 32-33;J. RAWLS, supra note 160, at 3-4; Tribe, supra note 130, at1374.

178. The focus on groups and the resulting submergence of the individualseem to threaten a return to the caste system or class orientation associated withthe feudal order, in opposition to which liberalism emerged originally. See supratext accompanying notes 162-63.

179. See Baker, Essay: Sandel on Rawls, 133 U. PA. L. REV. 895, 896-97, 905-07

15251986]

UCLA LAW REVIEW

the concept of group responsibility legitimates recognitionof the society or state itself-the largest group of all-not asan aggregate of sovereign individuals, but as an entity "witha life of its own distinct from and [potentially] superior tothat of all its members ..... ,s0 This contradicts the conceptof liberalism, which intends to prevent this very possibility.If the state has an independent existence, it may have an in-dependent definition of the good that can be imposed justifi-ably on citizens regardless of individual desires.Consequently, group responsibility threatens to foster the"collectivist dangers" of not only majority tyranny but alsostate totalitarianism.18'

The two objections of liberalism to group responsibil-ity-its blatant submergence of the individual and its poten-tial legitimation of state collectivism-are the principalreasons for the judicial and scholarly resistance to groupcausation. These objections explain why the courts havebeen largely unpersuaded by the deterrence and compensa-tion arguments in favor of group causation. Both of thesearguments are based on the social welfarist premise that so-ciety should seek to maximize the collective welfare of itsmembers-a premise adopted by the utilitarian theory oftorts.18 2 The social welfare theory sanctions both submer-gence of the individual and collectivism because it regardssociety in the aggregate as supreme. As a result, welfaristsand utilitarian torts theorists support group causation on de-terrence and compensation grounds, even though groupcausation disregards individual rights and responsibility. Incontrast, liberalism proclaims the individual's fundamental

(1985). Baker believes that Sandel's "account of the group subject could provide

the basis for a dangerous and unwarranted notion of group or community rights."Id. at 897.

180. J. RAWLS, supra note 160, at 264. Of course, Rawls uses the quoted lan-guage to reject just such a conception of society. I quote him here to emphasizehow completely group responsibility contradicts the central premises ofliberalism.

181. See Rosenberg, supra note 3, at 909.

182. See Wright, supra note 3, at 436-37; see also M. SANDEL, spra note 15, at3-7. Deterrence maximizes collective welfare by preventing avoidable accidentsor by reducing the level of unduly dangerous activities altogether. Compensationmaximizes welfare by alleviating accidentally caused suffering and the related

costs of dislocation as fully and quickly as possible. Neither of these approachesrequires that every individual's rights be respected as long as the net effect is tomaximize aggregate welfare.

1526 [Vol. 33:1473

BETWEEN TWO WORLDS

importance-the value adopted by the corrective justice the-ory of torts-and therefore holds that the good of the major-ity must never be pursued to the detriment of individuals'rights.183

The larger tension and accommodation between thewelfarist and liberal views has in effect bifurcated tort law.The result is a parallel accommodation between the utilita-rian and corrective justice theories that exists in practiceeven if it is not really philosophically coherent. The socialwelfarist view has supported the adoption of utilitarian anal-ysis on certain torts issues, most notably the negligence is-sue;' 84 but the liberal theory continues to commandadherence to corrective justice principles in the causationarea. 185 This adherence is especially steadfast because ifcourts adopted utilitarian arguments on the causation issueas well, they would upset the balance between liberal andwelfarist values. In effect, they would wholly displace the in-dividualist premise of liberalism from the law of torts, and,by implication, fundamentally challenge liberalism in its en-tirety. As a result, even if welfare-based deterrence or com-pensation arguments support the adoption of groupcausation rules, the need to maintain a firm place in tort lawfor the individualist premise of liberalism precludes such amove. The general rejection of group causation manifestsnot only an adherence to corrective justice rather than utili-tarian principles in the law of causation, but also a commit-ment at a deeper level to preserving the liberal theory ofsociety.' 86

183. See supra note 177.184. See Fletcher, supra note 3, at 556-57 & n.74, 563-69:

The question posed by the conflict [between strict liability and liabil-ity based on fault in the nineteenth century] was whether traditionalnotions of individual autonomy would survive increasing concern forpublic welfare. If the courts of the time had clearly perceived andstated the issue, they would have been shaken by its proportions.

Id. at 566. Courts today do recognize that this same issue underlies the groupcausation controversy and take it quite seriously.

185. See Wright, supra note 3; Wright, supra note 2, at 1822-26.186. The power of the liberal theory to limit the welfarist, utilitarian rationale

is understandable. Present day utilitarians argue primarily in terms of economicanalysis and focus on group conduct. See supra note 107. Despite their groupfocus, however, the economic theory of the marketplace on which utilitarians relyoriginated as a form of pure liberalism that held the individual economic actorsovereign and inviolate. No one could question, or even penetrate, the individ-ual's process of choice; in fact, any attempt to do so would be illegitimate. This

1986] 1527

UCLA LA W REVIEW

Some might argue that there really is no conflict be-tween this underlying commitment and group causation.Since the application of group causation will be limited es-sentially to mass and toxic torts cases involving corporatebusiness defendants who inflict untraceable harms, it willhave little impact on individuals per se and on individualfreedom. This is a disingenuous argument, and it has clearlyfailed to convince most courts. Apart from the observationthat individuals can and do own businesses that inflict un-traceable harms, scholars have already suggested extensionsof the group causation decisions that would expose ordinaryindividuals to group responsibility across the board.1 87 Ifgroup responsibility were accepted, its application ultimatelywould be extended to ordinary individuals; the distinctionbetween mass torts and sporadic accidents simply would nothold. Consequently, the threat to liberal individualism cannot be dismissed in this way.

The focus on mass torts, however, is important in an-other way. As these cases increasingly swell the ranks of tortplaintiffs with victims of indeterminately caused injury, whatwas once a tragic but limited problem becomes a tragic andmassive problem. Ultimately, this growing trend will se-verely test the courts' commitment to the preservation of lib-eralism. Although it is not difficult to justify rejecting Sindell-type claims on traditional causation grounds, this course ofaction offers scant satisfaction when growing numbers of se-riously injured victims are left without relief while plainlynegligent actors stand out of reach behind the protectiveshield of presumed causal innocence. Therefore, despite allthe resistance to accepting and extending the Sindell rule, thepossibility exists that it gradually will find wider acceptance.Nevertheless, the spread of group causation, as presentlyjustified by its advocates, would be just as troubling as itsrejection. The deterrence and compensation rationales thatare used to justify plaintiff recovery despite indeterminatecausation entirely disregard the individual responsibility

philosophical premise underlies both the economic concept of efficiency, orPareto optimality, as a social choice criterion, and the use of markets as a socialchoice process. Thus, while present-day economic analysis of tort law may departfrom liberalism's focus on the individual, the underlying economic theoryoriginates as part and parcel of liberalism. Consequently, liberal individualismhas the potential to override the economic theory.

187. See supra notes 90-95 and accompanying text.

1528 [Vol. 33:1473

BETWEEN TWO WORLDS

principle and the political theory of individual dignity andautonomy on which it rests. To adopt group causation onthese grounds is to invite, if not to initiate, a collectivist legalregime for corporate and individual actors alike.

The group causation cases and proposals thus bring tothe surface the longstanding tension between the underlyingpolitical theories of liberalism and social welfarism. At thesame time, they reveal that neither of these theories, orworld views, is ultimately capable of providing a satisfactorysolution to the problem of indeterminate causation. Thelegal battle over group causation in tort law thus appears tobe a battle between two worlds, neither of which is trulyhabitable.

III. A JUSTIFICATION FOR GROUP RESPONSIBILITY: THE

COMMUNITARIAN THEORY OF SELF AND SOCIETY

The foregoing parts of this Article traced the recent de-velopment of the concept of group causation in certain courtdecisions and commentaries, documented the widespreadresistance to this development, and located the reason forthe opposition in the ongoing tension between the politicaltheories of social welfarism and liberalism that underlie, andcompete for sway over, the legal system. The legal conflictover group causation appears to be a manifestation of alarger conflict of rival political visions, neither of which issatisfactory in light of the indeterminate causation problem.Thus, we are confronted with a kind of Hobson's choice be-tween two equally undesirable approaches. On one level ofanalysis, this conclusion is accurate. On another level, how-ever, it is possible to give a different and more optimisticinterpretation of what the group causation phenomenonrepresents. Rather than a simple assertion of welfarist overliberal values in the causation context, I suggest that thegroup causation decisions represent the courts' rough intui-tion that there is another world view-an alternative politicalvision distinct from both liberalism and social welfarism-according to which we can successfully resolve the indeter-minate causation problem and thereby avoid this Hobson'schoice.

Significant new work in political theory and in otherfields has begun to articulate an alternative framework as aresponse to the inadequacies of both liberal and social wel-

19861 1529

1530 UCLA LAWRE VIEW [Vol. 33:1473

fare theory. The "communitarian theory" of society claimsto avoid both the individualist excesses of liberalism and thecollectivist excesses of welfarism by stressing the central roleof community as a personal and social good and as a mediat-ing structure between the individual and the society as awhole.' 88 In the kind of intuitive leap that often occurs inthe common-law process, 89 courts introducing the groupcausation concept have, I believe, moved toward this newframework. This is not to say that the group causationcourts are familiar with the communitarian theory and con-sciously adopt it. On the contrary, lacking a clear under-standing or even an explicit awareness of this theory, theyhave continued to rely on the language of the old socialwelfarist framework, despite. its unacceptable threat of col-lectivism. Nevertheless, regardless of the language used, thecourts are intuitively, if unconsciously, responding to anddeveloping new ideas about community that are alreadymanifest in a wide variety of social, political, and intellectualcontexts, even if they have not yet been articulated clearly inlaw. ' 90

188. SeeJ. AUERBACH, JUSTICE WITHOUT LAW 4-6 (1983); M. SANDEL, supra note15, at 147-54, 172-74; Sandel, supra note 16, at 17;see also R. BELLAH, R. MADSEN,

W. SULLIVAN, A. SWIDLER & S. TipTON, HABITS OF THE HEART (1985) [hereinaftercited as R. BELLAH]; A. MACINTYRE, supra note 163; Macneil, supra note 164, at934-39; Murphy, supra note 16.

189. In his early work analyzing the courts' use of economic logic in developingnegligence doctrine, Posner adverted to the same kind ofjudicial intuition: In his

formulation of the negligence standard "[Judge Learned] Hand was adumbrating,perhaps unwittingly, an economic meaning of negligence." Posner, supra note 3, at32 (emphasis added). Hand's formulation "never purported to be original butwas an attempt to make explicit the standard that the courts had long [implicitly]applied." Id. Elsewhere, Posner realized that some may question whether judgesactually think in economic terms and clarified his point:

The reader may wonder whether it is plausible to attribute economicinsight to common law judges . . . . In searching for a reasonably

objective and impartial standard, . . . the judge can hardly fail toconsider whether a loss was the product of wasteful, uneconomicalresource use. In a culture of scarcity, this is an urgent, an ines-capable question. And in most cases, at least an approximation tothe answer is readily accessible to intuition and comnion sense.

R. POSNER, supra note 107, at 180-81 (emphasis added). Outside of tort law,scholars have argued that judicial decisions imposing new forms of contractualliability have resulted from "judicial intuitions" of structural changes in, and in-creased concentrations of, political and economic power in postindustrialAmerica. See Tobriner & Grodin, supra note 97, at 1247-49, 1266, 1282-83.

190. By "intuitively" I mean that courts are reaching conclusions or insights byunarticulated perception rather than by serial logical reasoning in which all the

BETWEEN TWO WORLDS

The shift to group causation is not simply the latestchapter in the ongoing rivalry between liberal and welfaristtheory in law. On a different level, it both reflects andmanifests a major shift to a new political vision that tran-scends the old conflict. Political theorists have begun to ar-ticulate this vision clearly. Now courts are beginning tointuit the new framework in the group causation cases. Thisalternative vision arises in response to changed perceptionsabout the real nature of human relationships, between injur-ers and victims and in general. This Article suggests thatthis vision represents a sounder and truer conception of so-ciety and should be welcomed in law as it is in politicaltheory.

Part III of the Article describes the communitarian cri-tique of liberalism, summarizes the basics of the communi-tarian theory and its advantages over liberalism, shows howthe group causation theories can be explicitly justified and atthe same time properly restricted by communitarian theory,and argues that the courts should pursue their initial intui-tion and develop the group causation rules. However,courts should develop these rules only on the explicit basisof communitarian reasoning, because this basis alone canboth justify group causation and yet prevent it from degen-erating into collectivism. The first portions of this Part focuson the comparison of the communitarian to the liberal the-ory. 19 ' The latter portions focus on the distinction between

premises of the conclusion are identified. Intuition can be either a conscious or

an unconscious process. In the case of the group causation decisions, the courts

are employing communitarian concepts intuitively, without clearly articulating thepremises behind these concepts. Most likely, they are also doing this uncon-sciously, without any direct awareness of the communitarian framework. Perhaps

the courts' actions are a response to ideas about community that are circulating in

other fields and in our general culture. See infra notes 262-67 and accompanyingtext. This Article argues that while the communitarian intuition found in these

cases, conscious or unconscious, is sound, it must not remain an intuition but

must be articulated in clear and logical terms. See infra text accompanying notes295-301. One scholar has actually suggested that the courts' intuitive movement

toward the communitarian vision can be seen in many fields, including contracts,torts, divorce law, constitutional law, and finance. See Presser, supra note 163, at

891-97. However, most of his examples seem more reflective of a move toward

social welfarism than a move toward communitarianism. See infra note 194.

191. This comparison is stressed because liberalism presently is the dominantframework and because the communitarian objections to the collectivist premises

of social welfare theory are quite similar to those of the liberal theory, which arewell-known.

15311986]

1532 UCLA LAW REVIEW [Vol. 33:1473

communitarianism and welfarist collectivism, a distinctioncrucial both to the political theory and to the justification ofgroup causation rules in tort law.

A. The Communitarian Critique of Liberal Individualism: TheStunted Self and the Fragile Community

The communitarian political theory shares liberalism'saversion to collectivism; therefore, it rejects social welfarismbecause that theory "fails to take seriously the distinctionbetween persons" and thus submerges the individual.192However, the communitarian theory finds liberalism equallyunsatisfactory because it "fails to take seriously our com-monality."1 93 Professor Michael Sandel is a major contribu-tor to communitarian political philosophy, and his views arerepresentative of a number of communitarian theorists. 9 4

192. M. SANDEL, supra note 15, at 140; see also id. at 144-45, 173-74.193. Id. at 174.194. See R. BELLAH, supra note 188; A. MACINTYRE, supra note 163; Murphy,

supra note 16. Some legal scholars have written commentaries that, to a greater orlesser degree, take a communitarian perspective without always articulating it assuch. See, e.g., Abel, A Critique of American Tort Law, 8 BRIT. J.L. & Soc'Y 199,205-08 (1981); Fletcher, supra note 3, at 547-49 & n.46; Frug, The City as a LegalConcept, 93 HARV. L. REV. 1057 (1980); Hutchinson, supra note 3, at 756, 766-71;Kronman, Alexander Bichel's Philosophy of Prudence, 94 YALE L.J. 1567, 1590-99(1985); Lindgren, Social Theory and Judicial Choice: Damages and Federal Statutes, 28BUFFALO L. REV. 711, 749-53 (1979); Macneil, supra note 164, at 934-48; Presser,supra note 163, at 891-99; Radin, Property and Personhood, 34 STAN. L. REV. 957,958, 971-78, 1011-15 (1982); Riskin, Mediation and Lawyers, 43 OHIo ST. L.J. 29,29-30, 56-57 (1982). However, it is important to distinguish scholars like Mac-neil, Radin, and Kronman, who use communitarian concepts explicitly and care-fully, from scholars like Abel, whose work contains insights similar to those ofcommunitarian theory but whose conceptual framework and proposals are clearlyrooted in social welfarism, a theory which the communitarian theory rejects asstrongly as it rejects liberalism. See Abel, supra, at 208-11; see also supra text accom-panying note 192; infra text accompanying notes 274-79. Other scholars fall be-tween these two poles. While some of the foregoing legal scholars appear to havebeen involved with, or influenced by, the Critical Legal Studies movement, theirwork presents anything but a "nihilist" approach-a term which some have usedto characterize the Critical Legal Studies movement. See Carrington, Of Law andthe River, 34 J. LEGAL EDUC. 222 (1984); Note, "Of Law and the River, ' and of Nihil-ism and Academic Freedom, 35J. LEGAL EDUC. 1 (1985). On the contrary, their workappears deeply committed to ensuring that legal rules express and effectuate tran-scendent values of human dignity. The same is true of the communitarian theoryas it applies to law.

At the same time, there are some important differences between communitar-ian legal scholarship and Critical Legal Studies. Some have argued that CriticalLegal Studies, while it challenges liberalism and associated legal concepts, haseither failed altogether to articulate clear alternative concepts and proposals orhas presented alternatives that are essentially socialist rather than communitarian

1986] BETWEEN TWO WORLDS 1533

Sandel starts by observing that liberalism as a theory of soci-ety is founded on certain assumptions about the nature ofthe self. He proceeds to criticize the theory by challengingthese assumptions. The concept of the self at the basis ofliberal theory is what Sandel describes as the detached, orantecedent, self.195 This self is separate from all others andthus "individuated" or unique; 96 but it is also separatedfrom its own circumstances so that it can be a totally freeactor or subject, a "sovereign agent of choice."'' 97 This con-ception is essential to liberalism because it expresses the to-tal liberation of the self from all deterministic constraints. 98

Such unconstrained liberty (consistent with the same forothers) was liberalism's antidote to the static society againstwhich it revolted. 99 For such a liberated self, the society ofliberal theory provided the environment necessary for self-fulfillment through sovereign individual choice.

The communitarian critique rests not only on the practi-

in character. (For my distinction between these two outlooks, see infra text accom-panying notes 239-45, 268-79.) See, e.g., Levinson, supra note 15, at 1469,1482-84; Presser, supra note 163, at 883 & n.60; Macneil, supra note 164, at919-29.

Even self-described communitarian legal scholars like Presser and Macneilhave had difficulty translating the communitarian vision into a practical guide forlegal policy and institutional reform. Presser's "alternative view," despite his lan-guage of community, is full of examples that seem more concerned with socialwelfare and social justice than with community. See Presser, supra note 163, at893-97. His language sometimes suggests the same. See, e.g., id. at 894 ("Thisright to one's job has been a principle of socialist law. ") (emphasis added); id. at895 ("[W]e have seen a marked retrenchment in rights accorded to criminal de-fendants as the rights of ... society generally seem to weigh more") (emphasis ad-ded). Macneil's articulation of the "community vision" is clearer than practicallyany other legal writer. I discovered his work after this Article was written, and wasdelighted to find what appears, in many respects, a kindred spirit. At the sametime, his translation of the communitarian vision into concrete terms of legal pol-icy is sketchy, focusing on proposals that would reduce large bureaucratic struc-tures and giving less attention to legal means for positively encouragingcommunity. See Macneil, supra note 164, at 942-43, 947-48. One major purposeof this Article is to show how the communitarian vision can be furthered throughpositive, community-strengthening legal rules and policies.

195. M. SANDEL, supra note 15, at 10.196. Id. at 50-55.197. Id. at 22. Such a self is "a creature whose ends are chosen rather than

given, who comes by his aims and purposes by acts of will, as opposed ... to actsof cognition." Id.

198. Thus, whatever the individual chooses is truly "his own path," id. at 60(quoting Nagel, Rawls on Justice, in READING RAWLS 9-10 (N. Daniels ed. 1973)),and not the path dictated by any outside conditions.

199. See supra notes 161-66 and accompanying text.

UCLA LA W REVIEW

cal difficulties of achieving this liberal vision, but also on thevarious conceptual flaws in the vision itself.20 0 Despite itsrhetoric of liberation, the liberal concept of the self is so re-stricted that it strips the self of important elements of iden-tity. Ironically, the ultimate result of this is theimpoverishment and disempowerment of the individualrather than his fulfillment. This in turn exposes the individ-ual to oppression by powerful groups and by the state. Inthe liberal vision, the concept of the self is "shorn of empiri-cally identifiable characteristics," including social roles andsociological circumstances, in order to liberate it from anypredetermined limits that would interfere with its exercise offree choice. 20 1 This vision fails because the self that remainsis a truncated, impoverished self. Liberalism "makes the in-dividual inviolable only by making him invisible, and callsinto question the dignity and autonomy [it] seeks above allto secure." 202

Dignity is undermined because the liberal self ultimatelyis not liberated but empty. 203 Devoid of commitments otherthan at a superficial level, since deeper commitments wouldthreaten to confine the self and constrain its total freedom,the self is left not "an ideally free and rational agent, but...a person wholly without character, without moral depth."2 0 4

Liberalism produces neither dignity nor liberation, but, inMurphy's words, "isolated individuals in various stages of adebilitating narcissism." 20

5 In effect, the refusal to recog-nize the possibility of deep commitment to any external so-cial reality results in a self that is totally free, but only at theprice of being stunted, because full development of the selfrequires just such commitment. 20 6

Autonomy is threatened in two ways. First, since everypure self is free of identifying and limiting characteristics, itis also devoid of intrinsic merit. As a result, society as awhole can override individual expectations and can even useindividuals as tools for social purposes. This is how Sandel

200. M. SANDEL, supra note 15, at 1.201. Id. at 95; see also id. at 93-94.202. Id. at 95.203. See R. BELLAH, supra note 188, at 80, 152.204. M. SANDEL, supra note 15, at 179.205. Murphy, supra note 16, at 155.206. R. BELLAH, supra note 188, at 81.

1534 [Vol. 33:1473

1986] BETWEEN TWO WORLDS 1535

characterizes liberalism's justification for affirmative ac-tion.20 7 Second, by definition, the stripped self is isolatedand vulnerable to external oppression. When individuals ac-cept the liberal concept of self, they withdraw into varyingdegrees of isolation and lose the protective elements of asso-ciation. 20 8 Ordinary individuals abandon their group tiesand are left alone, without any protective network of allies.As a consequence, they are highly vulnerable, both to corpo-rate economic interests and to the vastly increased machin-ery of the modern state. 20 9 The self of liberalism is thus"less liberated than disempowered" 210 and winds up lackingcoherence in the face of the external world. 211

207. "Once the self... is dispossessed, the claims of the individual fade tobetray an underlying utilitarianism which ... fails to take seriously the distinctionbetween persons." M. SANDEL, supra note 15, at 140. Sandel hypothesizes lettersof acceptance and rejection, written to convey the moral basis of affirmative actionpolicy according to liberalism. Paradoxically, this turns out to be the principle ofusing individuals and their attributes as instruments of wider social purposes.Thus, the rejection letter reassures the unsuccessful candidate that he was re-jected not because of any lack of personal merit, but simply because he did notmeet society's current needs. The successful applicant is congratulated, not be-cause of any personal or intrinsic merit, but simply because he was "lucky to havecome along with the right traits at the right moment" for society's purpose. Id. at142. Both individuals are told that they are intrinsically worthless and that theirtreatment stems from society's needs alone. Sandel does not reject affirmativeaction itself: he rejects the liberal justification for it. The policy itself might actu-ally be justifiable under the communitarian theory. See id. at 143-45.

208. In his observations of early American culture, de Tocqueville saw its indi-vidualism as a potential threat to individual freedom. See A. DE TOCQUEVILLE, 2DEMOCRACY IN AMERICA 107-13 (P. Bradley ed. 1945). It is one of the bitter iro-nies of liberalism that the sweeping rejection of class and group identifications asoppressive has exposed the individual to more oppressive forces than ever existedin premodern times.

209. See R. BELLAH, supra note 188, at 286. It can only be considered a crueljoke that individuals were persuaded to abandon important sources of socialpower, whether real or potential, for formal legal rights that have often proven tohave little value without the social power to claim and enforce them. See Bush,Dispute Resolution Alternatives and the Goals of Civil Justice." Jurisdictional Principles forProcess Choice, 1984 Wis. L. REV. 893, 982-84; Galanter, Why The "Haves" Come OutAhead. Speculations On The Limits of Legal Change, 9 LAw & Soc'v REV. 95, 137-39(1974); Rubenstein, Procedural Due Process and the Limits of the Adversary System, 11HARV. C.R.-C.L. L. REV. 48, 66-70 (1976).

210. M. SANDEL, supra note 15, at 178.211. A second aspect of the communitarian critique is that the liberal theory of

the self is simply at odds with our experience of reality. Individuals are not, anddo not perceive themselves to be, totally separate beings devoid of deep connec-tions to the world beyond them. See M. SANDEL, supra note 15, at 143, 179; see alsoR. BELLAH, supra note 188, at 152-62. Therefore, the pure liberal theory fails tohelp us order our social life. Despite its power, it remains at some level not be-

1536 UCLA LAW REVIEW [Vol. 33:1473

A corollary of the critique of the liberal concept of self isthat the liberal concept of community is likewise inadequate.The "fragile community" of liberalism is an aggregate ofconvenience, "put together to meet the utilitarian and ex-pressive needs of individuals." 212 But there is no recogni-tion that community is an absolute necessity upon which theindividual depends for self-protection and self-fulfillment.There can be no such recognition in liberalism because thiswould contradict the liberal premise that there is no realityof community co-equal with the reality of the individualself.2'3 At the extreme, by disregarding community, individ-ualism is carried so far that social cohesion is undermined.Ironically, this creates an environment in which individualfulfillment is itself impossible.214

Liberalism fails as a theory of society in the view of

lievable and leaves us simply disoriented and without any coherent world view.See id. at 163.

This criticism is related to the sociological criticism of liberalism, which ar-gues that empirical sociological observation demonstrates the reality and depth ofgroup influences and attachments for individuals. See Murphy, supra note 16, at141-42. Individuals are "embedded" in a social milieu and cannot be understoodor understand themselves without reference to it. However, the communitariancritique is not empirical but conceptual. It argues that apart from empirical evi-dence, the liberal idea of the self is not coherent as a concept. As a result, "wecannot coherently regard ourselves as the kind of beings the [liberal] ethic [ofKant or Rawls] ... requires us to be." M. SANDEL, supra note 15, at 14; see also A.MACINTYRE, supra note 163, at 241. Not only does the liberal self fail to cohere inthe face of the external world, it also fails to cohere with our own internal reality.

A third criticism is that the liberal theory is also internally incoherent. Focus-ing on Rawls' theory ofjustice as the most refined development of liberal theory,Sandel argues that Rawls' own definition of "justice as fairness," see M. SANDEL,supra note 15, at 15 (quotingJ. RAWLS, supra note 160, at 586) cannot be acceptedif the individual is viewed as a separate and sovereign self. The reason is thatRawls' basic fairness principle, the "difference principle," has redistributional ef-fects that inevitably violate individual sovereignty. See M. SANDEL, supra note 15, at70-71, 77-78. Sandel suggests that Rawls himself resolves this problem by im-plicitly altering his concept of the self, abandoning the detached self he originallyposits and recognizing an intersubjective self with deeply rooted attachments toothers. Id. at 79-82. Such a self can be told to share the fruits of its own achieve-ments with others without being violated because it experiences and directly par-ticipates in the resulting common good that unites the self with others. Id. at80-81, 143-44. The point of the communitarian critique is that the liberal con-cept of self proves insufficient even for liberalism, and the introduction of a differ-ent and sufficient concept results in inconsistencies between the theory and itsown basic premises.

212. R. BELLAH, supra note 188, at 50.213. See id. at 143, 334.214. See Murphy, supra note 16, at 154-55; R. BELLAH, supra note 188, at

275-83.

BETWEEN TWO WORLDS

Sandel and others because it is based on flawed conceptionsof the self and community that are ultimately incoherent andself-contradictory: "Liberalism cannot ... [continue] to in-sist upon the supremacy of the separate self. It must articu-late a new conception of the person . ,,215 It is this newconcept of the person that the communitarian theorypresents.

B. The Communitarian Theory: The Expanded Self and theConstitutive Community

Liberalism views the self as separate and detached fromothers and views community as constituted by separate individ-uals. Communitarian theory views the self as joined and at-tached to others and views community as constitutive of anexpanded self.2 16 Central to this theory is the role of com-munity as a personal and social good.

In contrast to the detached self of liberalism, com-munitarians conceive of the self as attached to othersthrough commitments so deep that the individual cannot un-derstand himself without reference to these commitments.Such a conception requires what Sandel refers to as an "in-tersubjective" form of self-understanding: "Intersubjectiveconceptions allow that the relevant description of the selfmay embrace more than a single, individual human being, aswhen we attribute responsibility or affirm an obligation to afamily or community or class or nation rather than to someparticular human being." 217 Given this expanded concep-'tion, "the self has to find its moral identity in and through its

215. Murphy, supra note 16, at 158.216. The terminology used here is borrowed from Sandel and has special

meaning in communitarian thought. See M. SANDEL, supra note 15, at 62-65,147-54. A community "constituted by" individuals is a structure made up, orformed by, individuals either to accomplish joint aims or to satisfy joint desires forinterrelationship. Self-aware individuals construct such a community to achievewhat they already know they want. A community "constitutive of" individuals is astrucure in which individuals discover what they want and who they are throughinteraction and shared reflection with others. Thus, the individuals are them-selves made up or formed by their participation in the community. Such a com-munity is vital to the individual's understanding and definition of'who he is in thefirst place. It is not merely an attribute or instrument of the self, but a constituentof the individual's identity. See id. at 150. The "constitutive" conception of com-munity stresses the active role played by community in the formation or constitu-tion of the self.

217. M. SANDEL, supra note 15, at 62-63.

19861 1537

UCLA LAW REVIEW

membership in communities such as those of the family, theneighbourhood, the city and the tribe .... " 218 The individ-ual "knows himself to be a participant in a form of lifethrough which his own identity is fulfilled." 21 9 Communitar-ian theory regards as crucial to self-understanding and ful-fillment the social attachments and commitments that liberaltheory rejects as constraining barriers to self-fulfillment. Inthe communitarian view, the individual can be understoodonly in relation to some community or communities withinwhich he finds his identity and personal fulfillment as amember. Thus, the concept of the self leads directly to theconcept of community.

Moreover, this is a very different kind of communityfrom the fragile aggregation of individuals found in liber-alism. This conception of community "penetrates the self":

[C]ommunity... describe[s] not just a feeling but a modeof self-understanding partly constitutive of the agent'sidentity .... [T]he members . . . conceive their identity... as defined to some extent by the community of whichthey are a part. For them, community describes not justwhat they have as fellow citizens, but also what they are...not merely an attribute but a constituent of theiridentity.220

According to this "constitutive conception" 22' of commu-

nity, participation in the community truly fulfills the individ-ual, not because it is pleasing in itself or because it helps himto get what he wants, but rather because it enables him todiscover and constitute his very self. Community is vital, be-cause the self is not a wholly antecedent being as liberal the-ory suggests, and it cannot be developed fully withoutparticipation in community. In mirror fashion, the conceptof self leads to that of community and the concept of com-munity leads to that of the self. Self and community,

218. A. MACINTYRE, supra note 163, at 205. Maclntyre calls this conception the"narrative view of the self," and he sees it as extending in time as well as in space,to prior and succeeding generations. Id. at 204-07; see also R. BELLAH, supra note188, at 153.

219. R. BELLAH, supra note 188, at 162; see also Murphy, supra note 16, at 156(explaining that the "crucial status is membership in a political society").

220. M. SANDEL, supra note 15, at 150 (emphasis in original). This constitutiveconception of community also differs from the sentimental conception, accordingto which a community is simply a consensual gathering of separate individualswho happen to share a certain preference for sociability and cooperation per se.See id. at 148-49.

221. Id. at 150.

[Vol. 33:14731538

1986] BETWEEN TWO WORLDS 1539

although distinct, are inextricably connected. 222

The main policy implication of the communitarian the-ory is the need to place more emphasis on the formation of"communities in the strong sense," communities capable ofnot merely procuring goods for the stunted self but consti-tuting the good of the expanded self.22 3 The benefits ofsuch "constitutive communities" would be numerous. First,constitutive communities would replenish the self, restore asense of dignity and substance to individuals, and relievealienation and emptiness.224 Second, constitutive communi-ties would help protect individuals from exploitation andoppression by joining them in strong networks of allies thatcould more effectively confront the corporate and bureau-cratic centers of power characteristic of modern society.2 25

In these ways, community formation would actually supportindividual self-fulfillment-the core value of liberalism.Third, the existence of constitutive communities would, byenabling groups to speak with more effective voices in publicforums, enrich debate on public issues and thus improve thequality of the democratic process and its resultant deci-sions.2 26 In this manner, community strength would further

222. In liberal theory, the self is separate not only from other selves, but alsofrom any community that is formed as an instrument of its constituents. Thus,community depends upon and is distinct from its constituents. Given this view,while it is possible to argue that community may have instrumental value in fur-thering individual fulfillment, it is difficult to see what value there is in communityper se. In communitarian theory, the intersubjective understanding of self linksindividuals and forms communities, which in turn act as the vehicle for furtherunderstanding and constituting the expanded self. Therefore, community is a"good" in the same sense and for the same reason that self-fulfillment is a"good": community life is self-fulfillment for the expanded self. See id. at 64-65;see also infra notes 223-30 and accompanying text.

223. In other words, participation in a constitutive community helps the indi-vidual discover and form his identity, including his definition of what is good.This definition, furthermore, necessarily takes others into account, for the ex-panded self has deep and undeniable connections to others.

224. See M. SANDEL, supra note 15, at 179-80; Macneil, supra note 164, at936-38.

225. See R. BELLAH, supra note 188, at 159-62, 286; M. SANDEL, SUpra note 15, at143-44. Much of the Critical Legal Studies work on community seems to focus onthis value. See supra note 194.

226. See Kronman, supra note 194, at 1591-96. Kronman sees a strong commu-nitarian strain in Professor Bickel's work, that recognizes the essential value of"intermediate communities" in American political life and argues for constitu-tional rules, such as allowance of "malapportionment in the service of group rep-resentation," that preserve and strengthen such communities. See id. at 1597.

UCLA LAW REVIEW

the achievement of the collective good-the primary value ofthe social welfare theory.

The strong community would become a crucial struc-ture mediating between the individual and the society as awhole. This mediating role, as implied in the foregoingpoints, would make possible the joint fulfillment of both lib-eral and welfarist aims, which increasingly have been per-ceived as mutually exclusive. This capacity to reconcileopposing liberal and welfarist goals would itself be a majorbenefit of increased attention to community.

All the benefits of community mentioned thus far areactually linked to the original liberalism-social welfarismframework. Community is valuable because it is an effectiveinstrument for attaining the goals of one or the other theory,or indeed of both. However, the constitutive communitiessought by communitarian theory would have additional ben-efits unrelated to the values of either liberalism or social wel-farism. 227 Membership in a constitutive community woulddemand constant and shared reflection by individuals on thenature of the common good. Such reflectiveness is a goodin itself because it fosters the capacity for reflection that is adefining quality of the human being. Moreover, shared re-flection on common purposes, and collaborative action toachieve them, both confront the individual with his ownboundedness and require that he transcend it at least in part.This draws forth the paradoxical capacity of the limited selfto achieve self-transcendence, which is another essentialquality of the human being.2 28 Constitutive communitiescan thus facilitate and demand realization of the highest ca-pacities of the self-a value quite different from that of self-fulfillment in the ordinary sense. 229

Finally, through the shared reflection and common en-terprise required in constitutive communities, individualscould more readily realize and sustain intrinsic values suchas friendship, loyalty, and charity. Such values do not derive

227. See Macneil, supra note 164, at 934-35 & n.131.228. See id. at 900 n.5, 935-36. Of all the legal writers I have found taking

some sort of communitarian point of view, only Macneil recognizes the self-tran-scendence value.

229. See id. at 935 n.134, 939-43. Macneil distinguishes this value from that of"establishment of 'ego identity,' " which he sees as underlying the Critical LegalStudies vision and distinguishing it from communitarianism. Id. at 940 (citingLevinson, supra note 15, at 1483).

1540 [Vol. 33:1473

BETWEEN TWO WORLDS

their true worth from the role they play in achieving individ-ual fulfillment or social welfare, and they do not vary accord-ing to circumstances. They are good in themselves,unqualified and universal. Revelation and realization ofthese absolute goods or "virtues" (or aspects of the "Abso-lute Good") would be a distinct and profound benefit ofconstitutive communities. 230

The foregoing summarizes the essence of the communi-tarian alternative to, or revision of, liberalism. Through itsfocus on the key role of community, the communitarian the-ory would maintain and even enhance the dignity of the indi-vidual person, while balancing and thus curing theindividualist excesses of liberalism. Although at odds withthe dominant liberal orthodoxy, the communitarian ap-proach claims a long and respectable pedigree, extendingback to the biblical tradition of religious community, 23 ' thecivic republican tradition of classical civilization and the ren-aissance,232 and the federalist and republican political theo-ries of the founding fathers. 233

Descending from the philosophical to the practicallevel, what specific implications does the communitarian the-ory have for law and social policy? To begin with, communi-tarian theory would modify liberalism's exclusive emphasison measures and rules that protect individual autonomy.Although such measures would still have an important placein a communitarian system, equal importance would attachto measures that increase the awareness of the intersubjec-tive dimensions of self and foster community formation andsolidarity.2 34 On this practical level, communitarian theoryhelps to explain the development of group causation and re-

230. See A. MACINTYRE, supra note 163, at 167-89, 204-05; M. SANDEL, supranote 15, at 179-83.

231. See R. BELLAH, supra note 188, at 28-31, 333; see also J. AUERBACH, supranote 188, at 19-30.

232. See R. BELLAH, supra note 188, at 30-31, 335; A. MACINTYRE, supra note163, at 220-22.

233. See R. BELLAH, supra note 188, at 30-31, 251-56; R. NEUHAUS, THE NAKEDPUBLIC SQUARE: RELIGION AND DEMOCRACY IN AMERICA (1984); Sandel, Book Re-view, NEW REPUBLIC, June 10, 1985, at 37, 39 (reviewing J. DIGGINS, THE LOSTSOUL OF AMERICAN POLITICS: VIRTUE, SELF-INTEREST, AND THE FOUNDATIONS OFMODERN LIBERALISM (1984)). See generally D. EPSTEIN, THE POLITICAL THEORY OF

THE FEDERALIST (1984).234. See R. BELLAH, supra note 188, at 162, 251, 259-63; Macneil, supra note

164, at 939.

19861 1541

UCLA LAW REVIEW

sponsibility theories in tort law, and it provides both ajustifi-cation of these theories and a critical understanding of theirlimits and dangers.

C. The Communitarian Justification for Group Responsibility:Addressing the Expanded Self and Encouraging Community

Without the communitarian framework, court decisionsthat adopt group responsibility can only be seen as an asser-tion of the primacy of welfarist values, embodied in deter-rence and compensation goals, over liberal concerns forindividual autonomy. 235 Yet even the group causationcourts admit they are uncomfortable with this "trade-off"and its collectivist implications. 236 I suggest that what in-spired these courts to fashion the group causation conceptwas not a higher commitment to welfarist theory, but ratheran intuition of an alternative vision according to whichgroup responsibility does not undermine individual dignity,but actually enhances it. This alternative vision is the com-munitarian theory, which supports group responsibility rulesin just this manner on two distinct levels. First, it displacesthe objections of liberalism that these rules threaten individ-ual autonomy. Second, it provides the positive justificationthat these rules facilitate expanded self-understanding andcommunity development, and thus ultimately enhance indi-vidual dignity.

In response to the liberal objection that group responsi-bility submerges the individual, communitarian theoryshows that this criticism is both ironic and false. Accordingto the liberal view, rules based on individual rights and re-sponsibilities are necessary to preserve the autonomous selfand to facilitate its exercise of free choice. However, com-munitarian theory demonstrates that the autonomous indi-vidual of liberalism is in reality stunted and disempowered.If so, it is unnecessary and even undesirable to maintain thestatus quo by adhering to the IR principle. Moreover, groupresponsibility does not undermine the individual; it simplyrecognizes and encourages the development of commu-nity. 23 7 The ultimate effect of this is to strengthen the indi-

235. See supra text accompanying notes 157-86.236. See supra text accompanying notes 98-110.237. See infra text accompanying notes 247-60.

1542 [Vol. 33:1473

BETWEEN TWO WORLDS

vidual by expanding and constituting the self,238 a result thatleads not to submergence but emergence of the individual.

The second liberal objection to group responsibility,that it threatens the individual by facilitating state collectiv-ism, reflects a misunderstanding of group responsibility thatarises from liberal theory itself. The liberal political uni-verse is a dualist realm that includes only the state and theindividual. 23 9 As Murphy comments, liberalism "wavers be-

tween anarchy and absolutism." 240 Consequently, the lib-

eral critic assumes that a shift from individual to groupresponsibility can only mean a shift to collective societal re-sponsibility and eventually state collectivism. Many propos-als for tort reform-for example, those involving state-administered general social insurance-actually would re-place the IR principle with societal responsibility. 241 How-ever, these proposals do not represent group responsibility,nor are they supported by communitarian theory. Rather,state collectivism is part and parcel of the dualist politicalworld view. It is the social welfarist mirror image of liberalindividualism, the other pole of the bipolar universe. Com-munitarian theory rejects the collectivist along with the indi-vidualist vision. At the same time, it clarifies that groupresponsibility actually resists this polarization by reintroduc-ing and legitimizing the intermediate groups that liberalismdiscredited at its inception. 242 Group responsibility, by fos-tering these intermediate groups,243 helps repopulate thepolitical universe and thus represents an alternative to and a

238. See R. BELLAH, supra note 188, at 152-53.

239. See supra text accompanying notes 161-66.

240. Murphy, supra note 16, at 145; see also id. at 144-46.

241. See T. ISON, THE FORENSIC LOTTERY: A CRITIQUE ON TORT LIABILITY

(1968); Abel, supra note 194; Englard, supra note 3, at 49-50; Franklin, supra note

95; Ison, Tort Liability and Social Insurance, 19 U. TORONTO L.J. 614 (1969); Pierce,

Encouraging Safety: The Limits of Tort Law and Government Regulation, 33 VAND. L.

REV. 1281 (1980). The pure "compensation" theory of tort law-the theory that

the system's goal should simply be to compensate all victims of injury, see, e.g.,

Note, Beyond Enterprise Liability, supra note 123, at 706-finds its justification, if it

has one, in a collectivist theory of society. See Englard, supra note 3, at 68-69. For

a definition of "social insurance," see supra note 106, which also distinguishes

social insurance from other insurance mechanisms that involve group rather thansocietal responsibility.

242. See R. BELLAH, supra note 188, at 162, 212.

243. See infra text accompanying notes 247-60. Presser notes that the Critical

Legal Studies scholars have focused on the reconstruction of "intermediate socie-

ties" as a central issue. See Presser, supra note 163, at 892 n.6.

15431986]

UCLA LA W REVIEW [Vol. 33:1473

safeguard against state collectivism.244 The collectivist ob-jection is fundamentally mistaken.2 45

The preceding arguments assume that group responsi-bility actually facilitates expanded self-understanding andencourages community formation. If so, the liberal objec-tions are displaced. More important, because communitar-ian theory shows how these effects enhance individualdignity, group responsibility is justified as ultimately enhanc-ing individual dignity, in addition to furthering other impor-tant values, such as fostering the self's capacity for reflectionand transcendence, and revealing and realizing universaland absolute goods or virtues. 246 However, the reader mayreasonably ask: Are the stated assumptions about the effectsof group responsibility valid? Group responsibility theoriesin tort law hold individuals responsible for each other's con-duct when all are members of a common group and allowindividuals to recover for each other's harm when all aremembers of a common group.2 47 I maintain that on several

244. Levinson implies that one reason Critical Legal Studies scholars have notthoroughly embraced the communitarian approach is that they fear the emphasison community may mean both submergence of the individual and collectivism.See Levinson, supra note 15, at 1480, 1483-84. If this is true, it is ironic, for itimplies that Critical Legal Studies scholars accept the liberal objections and arethus still wedded at some level to the liberal individualist framework they soharshly criticize. Macneil argues that this is largely the case. See Macneil, supranote 164, at 939-42.

245. A variation of this collectivist objection, which might be called the "tribal-ist objection," also fails to undermine communitarian theory. (I am indebted forthis suggestion to my colleague, Professor Lawrence Joseph.) In essence, this ob-jection rests upon the assumption that group responsibility empowers groups thatcan then oppress both individuals and less powerful groups. This took place inthe feudal period, and liberalism developed as a response to just such oppression.Nevertheless, this objection misses the point of group responsibility in communi-tarian theory. Unlike liberalism's abolition of intermediate groups, communitar-ian theory does not advocate eradication of individual consciousness. See R.BELLAH, supra note 188, at 275-80; M. SANDEL, supra note 15, at 143, 150. Rather,it seeks to reinstate the intermediate group as a mediating partner between theindividual and the state in the political universe. See R. BELLAH, supra note 188, at212. In the pluralist universe of communitarian theory, the group serves to pro-tect the individual from oppression by the state and other groups, while the statemediates between groups and protects against group oppression of individualmembers. The individual is further protected from oppression within any givengroup by his membership in other groups. This aspect of multiple and mobilemembership distinguishes modern communitarianism from its medieval anteced-ents. Likewise, group responsibility does not necessarily mean the elimination ofindividual responsibility. See infra notes 274-82 and accompanying text.

246. See supra text accompanying notes 227-30.247. See supra notes 22-97 and accompanying text.

1544

BETWEEN TWO WORLDS

levels, such measures encourage individual injurers andvictims to expand their perceptions of self, acknowledgetheir links with others and thereby form or strengthencommunities.

On the first level, when responsibility is placed ongroups or is viewed as owing to groups, the legal processitself is transformed from an individual to a group process.As a result, individuals form associations with one another topursue or defend claims. 248 The groups thus formed maycontinue to exist for considerable periods and develop intolasting communities, especially in the context of major litiga-tion. In the Agent Orange249 case, for example, Judge Wein-stein's principles for structuring the distribution of thesettlement funds recognized the way in which the litigationhad coalesced individual veterans into groups, and he sug-gested that the distribution plan should try to strengthenthis community formation process: "The task of organizingrepresentatives of. . .veterans into an effective group...must be given priority .... Veteran representatives shouldbe selected in a manner that will further unite the class ....Some form of National Center ... should be funded to pro-vide . . . veterans with a visible, central source of...power.'"250

On a second level, "[c]ollective responsibility not onlyexpresses [group] solidarity but also strengthens it."25

t Inother words, group responsibility creates incentives thathelp strengthen both expanded self-awareness and commu-nity. It conveys a direct message to injurers and victims that,for purposes of liability and recovery, their actions and cir-

248. See McGovern, supra note 96, at 8, 16 (reporting the increase in collabora-tive and joint efforts by both plaintiffs and defendants in toxic tort litigation, in-cluding formation of plaintiff and defense associations).

249. In re Agent Orange Prods. Liab. Litig., 597 F. Supp. 740 (E.D.N.Y. 1984).250. Id. at 858-59. On the injurer side, defendant manufacturers and insur-

ance companies in the ongoing litigation over asbestos-caused injuries have ac-knowledged their identity as a community. They have formed a permanent,jointly funded asbestos compensation facility to discharge their community obli-gation to injured individuals. See Maher, Asbestos Extravaganza, CAL. LAw., June1985, at 60, 64. While these examples are taken from the mass tort context, thesame argument could apply to the sporadic accident situation if group responsi-bility were more explicitly applied there. See supra text accompanying notes90-95. One would then expect to see the formation or strengthening of associa-tions of potential victims or injurers-homeowners, merchants, consumers, boat-ers, and the like-as categories of responsible actors were legally recognized.

251. J. FEINBERG, supra note 11, at 233; see also id at 238-39.

15451986]

UCLA LAW REVIEW

cumstances will be considered linked to those of others withwhom they form a cognizable group or community.2 52 Thismessage provides a strong incentive for individuals to beaware of their practical attachments to others and to takethose attachments seriously-to conduct themselves not asisolated individuals but with constant attention to the affairsof others with whom they are associated.

In practice, group causation would require an individualproducer, for example, to attend to the safety practices offellow producers making similar products and to expectthem to attend to his safety practices, because each stands tobe affected by what the others do or fail to do: "Manufactur-ers unable to prove that they could not have produced theinjury-causing product, will only benefit from safer behaviorif they can nudge the entire industry into making the sameor similar modifications to increase safety .. ,"253 Each in-dividual producer would have an incentive to be concernedabout and involved with others. The "leading producer"would have an incentive to share risk and safety informationand to control "marginal" producers. At the same time, the"marginal" producer would have an incentive to police the"leader." The safety risks that a large producer might toler-ate because of the relatively limited potential impact on himcould be catastrophic to the marginal producer who wouldhave to share the consequences. Ultimately, by inducing di-alogue and positive interaction among all the members of aninjurer group, group responsibility encourages injurercommunity. 254

252. For a discussion of how to define a "cognizable" group or community, seeinfra text accompanying notes 280-300.

253. Note, supra note 3, at 319. This commentator worries, however, that suchproducer interaction might violate antitrust prohibitions. See id. at 319-20. Thisis a valid concern under current legal theory, but producer communities may notbe viewed as an evil under the communitarian theory. Furthermore, producercommunities certainly exist, and they operate for some nonprohibited purposes,such as political lobbying. Refusing to foster producer-community responsibilitywhile permitting producer communities to seek political advantage makes littlesense even under liberalism. This is another example of how liberalism leaves theindividual more vulnerable by disregarding the reality of the existence of certainpowerful groups and thus allowing them to act unregulated. If antitrust theorywould impede the imposition of group responsibility on producers, then it tooshould be reexamined and revised.

254. This example is taken from an area in which group causation rules arepresently used, but group causation would create the same incentives in any cog-nizable group of potential injurers, whether in the mass or sporadic tort context.

[Vol. 33:14731546

BETWEEN TWO WORLDS

From the victim's perspective, group responsibility alsoprovides incentives for expanded self-understanding andcommunity formation. One major consequence of group re-sponsibility is that a victim often will recover less than hisfull damages, while others injured by unrelated causes willget a windfall.2 55 If, as Judge Weinstein conceded in theAgent Orange case, "[i]t is not possible for class members toreceive significant individual compensation from the settle-ment funds," 256 what then will the individual victim receive?In essence, the individual is asked to feel compensated bybenefits that go to others in the group. In his decision,Judge Weinstein sought to assure "acceptable targeting oflimited settlement funds," to fund "some form of NationalCenter for Vietnam Veteran assistance," to establish a pro-gram for "genetic counseling," and to aid "children of mem-bers of the class with birth defects." 257

Despite the reasonableness of these goals, one maywonder: What happens to the permanently disabled individ-ual who has suffered hundreds of thousands of dollars oflifelong losses yet receives only a small cash award, if any-thing? Furthermore, what if this individual never needs ge-netic counseling, childrens' aid, or any of the otherprograms designed to benefit "the class"? On the one hand,this result seems unfair and cruel to the individual victim.

This incentive for group self-policing is noted by Feinberg, who claims that it isnot as important when external controls are present and effective. See J. FEIN-BERG, supra note 11, at 234-41. Although Feinberg believes external controls areadequate in modern society, external policing actually becomes less practical andeffective in preventing injuries as technology becomes more complex. Inciden-tally, this point answers the argument that group responsibility would reduce de-terrence by shielding the careless individual from the consequences of his actions.See, e.g., Note, supra note 3, at 320-21. If anything, mutual policing would increasedeterrence.

255. See supra text accompanying notes 43-65.256. In re Agent Orange, 597 F. Supp. 740, 859 (E.D.N.Y. 1984).257. See id. at 858-61. Such class remedies have been advocated and some-

times applied in other cases. See Bigelow v. RKO Radio Pictures, 327 U.S. 251,264 (1946) and Pettway v. American Cast Iron Pipe Co., 494 F.2d 211, 260-63(5th Cir. 1974) (both involving class wide estimation of total damages); Eisen v.Carlisle &Jacquelin, 52 F.R.D. 253, 264 (S.D.N.Y. 1971), rev'd, 479 F.2d 1005 (2dCir. 1973), vacated and remanded on other grounds, 417 U.S. 156 (1974) and Bebchickv. Public Util. Comm'n, 318 F.2d 187, 203-04 (D.C. Cir.), cert. denied, 373 U.S. 913(1963) (both involving proposals for establishing a fund to benefit class membersgenerally); see also Delgado, supra note 12, at 901 n.7; Rosenberg, supra note 3, at916-24. See generally Developments in the Law-Class Actions, 89 HARV. L. REV.1318, 1524-36 (1976).

1986] 1547

UCLA LA W REVIEW

On the other hand, it encourages him-actually forces him-to view himself not as an individual victim at all, but as amember of a suffering group, in which his pain is shared byothers and eased by the alleviation of their pain. This asks agreat deal of the individual victim. It asks him to go beyondthe limits of his individual self altogether. Sandel writes ofsuch self-transcendence in a similar context:

Where [a] sense of participation in the achievements...of (certain) others engages the reflective self-understand-ings of the participants, we may come to regard ourselves• . . less as individuated subjects . .. and more as mem-bers of a wider.., subjectivity, less as 'others' and moreas participants in a common identity ....

One consequence of an enlarged self-understanding... is that when 'my'... life prospects are enlisted in theservice of a common endeavor, I am likely to experiencethis less as a case of being used for other's ends and moreas a way of contributing to ... a community I regard asmy own . . . and with which my identity is bound.2 58

If expanded self-understanding and a sense of commu-nity can be gained through a sense of common achievement,they can also be gained, if more poignantly, through a senseof common suffering on the part of individual members of avictim group. Furthermore, the expanded self-awarenessgenerated by group responsibility combines with the incen-tives of group recovery to encourage actual and potentialvictims of injury to solidify the links between them by form-ing actual associations or communities to defend their com-mon health and well-being.259

258. M. SANDEL, supra note 15, at 143.259. Even before the event of injury, recognition of group responsibility en-

courages individuals exposed to common risks as potential victims to realize theirconnection to one another and to associate in communities and act on that realiza-tion. Such responses of community formation by victims or potential victims havebeen evident in situations such as the Love Canal case, the Three Mile Islandnuclear reactor "incident," and the Dalkon Shield case. For a discussion of theLove Canal case, see L. GIBBS, LOVE CANAL: MY STORY (1981); Cary & Katz, Howto Track Down Toxins, NEWSWEEK, May 6, 1985, at 81; Shribman, A Woman Trans-formed by a Cause, N.Y. Times, Oct. 30, 1981, at A20, col. 4; Brown, Love CanalU.S.A., N.Y. Times, Jan. 21, 1979, § 6 (Magazine), at 23. For a discussion ofThree Mile Island, see Walsh, Resource Mobilization and Citizen Protest in CommunitiesAround Three Mile Island, 29 Soc. PROBS. 1 (1981); Witt, Vital Social Issues Reach HighCourt, Christian Sci. Monitor, Dec. 31, 1982, at 1, col. 4; Cummings, Leaders ofNuclear Protests See a Shift in Strategy, N.Y. Times, Sept. 21, 1981, at B10, col. 3. Fora discussion of the Dalkon Shield case, see S. ENGLEMAYER & R. WAGMAN, LORD'S

JUSTICE (1985). See also Menkel-Meadow, Toward Another View of Legal Negotiation:The Structure of Problem Solving, 31 UCLA L. REV. 754, 805 (1984). Menkel-Meadow

1548 [Vol. 33:1473

1986] BETWEEN TWO WORLDS 1549

Group responsibility in tort law thus furthers the com-munitarian goals of fostering expanded self-understandingand encouraging community both directly and indirectly. 260

As a result, communitaran theory not only displaces the ob-jections of liberalism, but also justifies group causation asultimately enhancing individual dignity as well as furtheringother important values. Although courts that have adoptedgroup causation may have intuited this communitarian vi-sion, they typically have justified their rulings by deterrenceand compensation rationales that are based on social welfar-ism. 261 Such rationales, with their collectivist implications,are plainly inadequate to convince the majority of courts thatgroup causation will not undermine the liberal, individualistvalues they hold preeminent. Nevertheless, courts believingin liberal theory should be willing to accept group responsi-bility if it rests explicitly on an articulated communitarian ra-tionale. For it would then be clear that group causationposes no threat to individual dignity, but rather helps tostrengthen the individual by following a communitarian vi-sion that moderates liberal individualism without embracingwelfarist collectivism. If, as I contend, most courts primarilyobject not to the results but rather to the collectivist ration-ale and implications of the group causation decisions, then

notes that "[i]n some recent products liability cases.., plaintiffs who could not befully returned to their previously uninjured selves have requested that the defend-ants rewrite their warnings in order, at least, to prevent similar injuries to others."While Menkel-Meadow attributes this to the "altruistic objectives" of the individu-als involved, it is also possible that this request reflects the kind of communityconsciousness that Sandel describes. See supra text accompanying note 258.

260. There is a more general level at which group responsibility in torts mighteffect communitarian goals apart from direct impact through group legal processand behavioral incentives. By encouraging actual group consciousness among in-jurers and victims, and even by simply recognizing concepts such as group injury,group conduct, and the like, the legal system legitimizes the concept of commu-nity in a powerful way. This may encourage individuals to recognize and experi-ence their attachments to others outside the context of tort injuries and to formassociations and communities on this basis. Applying communitarian concepts totort law in particular may be very influential. After all, the communitarian visionof individual self-transcendence tends to appear highly abstract and idealized. Isit in any way realistic to suppose that human beings can and do experience selfand others in the way communitarian theory suggests? Using an essentially com-munitarian concept to deal with the harshest and grimmest realities of an injurycase grounds this vision in reality in a uniquely powerful way. If we can acknowl-edge expanded conceptions of self and recognize community in these situations,this is a strong argument that the communitarian vision can be realized in othersituations as well.

261. See supra text accompanying notes 98-110.

UCLA LA W REVIEW

the thorough and explicit articulation of the communitarianintuition already implicit in these decisions should changethe minds of many courts.

Nevertheless, it remains true that communitarian theorysupports group responsibility only by departing from tradi-tional liberalism. Therefore, while the communitarian visionin effect mediates between the poles of liberal individualismand welfarist collectivism, it still presents us with a choicebetween two worlds, or world views. Although advocates ofthe communitarian view can be found in growing numbers,not only in political philosophy262 but also in sociology, 263

history, 264 psychology, 265 medicine, 266 and even law, 267 somecrucial questions remain. If the communitarian critiquehelps to illuminate the limits of liberalism, what are theunarticulated limits of the communitarian theory? Despiteall their flaws, are liberalism- and the IR principle still prefer-able to the communitarian theory, whose limits may be moreserious and problematic than first anticipated?

One question in particular goes to the heart of the is-sues raised in this Article and echoes the liberal objectionsto group responsibility: 268 Can any adequate definition orlimit be placed on the term "community," so that theboundaries between society, community, and individual arerecognizable? In the context of tort law in particular, canany adequate definition or limit be placed on the term''group,'' so that group responsibility neither diffuses intogeneral societal responsibility nor wholly eliminates individ-ual responsibility? If affirmative answers cannot be given tothese questions, then despite earlier arguments to the con-trary, 269 communitarian theory and group responsibilitywould appear to lead to collectivism, submergence of the in-dividual, or both. That is, communitarian theory would sim-

262. See A. MACINTYRE, supra note 163; M. SANDEL, supra note 15; Kronman,supra note 194; Murphy, supra note 16.

263. See R. BELLAH, supra note 188.264. SeeJ. AUERBACH, supra note 188; Sandel, supra note 233.265. See C. GILLIGAN, IN A DIFFERENT VOICE: PSYCHOLOGICAL THEORY AND WO-

MEN'S DEVELOPMENT (1982).266. SeeJ. LYNCH, THE LANGUAGE OF THE HEART (1985).267. See Kronman, supra note 194; Macneil, supra note 164; Radin, supra note

194; see also Hutchinson, supra note 3; Lindgren, supra note 194. But see supra note194.

268. See supra text accompanying notes 173-82.269. See supra text accompanying notes 236-47.

[Vol. 33:14731550

BETWEEN TWO WORLDS

ply be social welfarism in different clothes. Communitariantheory helps to justify group responsibility. Does it simulta-neously justify societal responsibility and eliminate individ-ual responsibility? The links established in this Articlebetween communitarian theory and group causation in tortlaw are quite important in this context, because the concreteworld of injury and compensation provides at least the be-ginning of an answer to these final questions.

D. The Limits of the Communitarian Justification: RecognizingGroups and Respecting Individuals

Although communitarian political theory has greatly il-luminated the limits of liberal concepts, it has not beennearly as clear about defining its own primary terms. Thedefinition of "community" is itself fraught with vaguenessand confusion in the literature, which threatens to merge"community" with society on the one hand, and with the in-dividual on the other.270 Lack of clarity is not necessarilyfatal to political theory, which is abstract by nature and maybe less concerned with defining community than with deem-phasizing the centrality of the individual. However, in thecontext of concrete policies and legal doctrines, lack of rigorand clarity is a serious problem.

The group causation theories are particularly suscepti-ble to a vagueness problem because without adequate defini-tions, group responsibility diffuses into general societalresponsibility and wholly eliminates individual responsibil-ity. Proposals for "group responsibility" in torts have suf-

270. Some political theorists draw no clear boundary between community andsociety. Murphy defines community as a "political society," by which he means "awhole people." See Murphy, supra note 16, at 146. Bellah implies that "the na-tional society" can be defined as the essential community. See R. BELLAH, supranote 188, at 250-51, 285. On the other hand, Sandel criticizes the "societal" defi-nition of community and argues for a more "local or particular" definition. See M.SANDEL, supra note 15, at 145-46; see also A. MACINTYRE, supra note 163, at 244-45.

However, Sandel's own terminology leaves it very unclear how he would definethe other crucial boundary, that between community and individual. See M.SANDEL, supra note 15, at 62-63, 143, 150, 173; see also A. MACINTYRE, supra note

163, at 204-07. It is clear that Sandel thinks there is such a boundary and that hisconcept of community is not intended to absorb wholly the individual. See M.SANDEL, supra note 15, at 173 (insisting that the individual is "never finally dis-solved"); see also id. at 150. It is not as clear that MacIntyre believes in this kind ofboundary. See A. MACINTYRE, supra note 163, at 205. However, without a definedboundary, absorption of the individual is impossible to prevent.

15511986]

UCLA LA W REVIEW

fered from precisely this confusion. Some proposals haveadvocated nonjudicial compensation schemes funded bygeneral tax revenues. 27I Essentially, these are social insur-ance plans that would make society at large responsible forall accidental harm. Other proposals, like Robinson's "riskcontribution liability," 272 would retain individual tort actionsbut would transform every individual situation into one ofgroup responsibility. By imposing liability and awarding re-covery only on probabilistic grounds, these proposals wouldeliminate individual responsibility even in a case of unequiv-ocal actual causation of harm.273 Neither of these ap-proaches represents a form of group responsibility that canbe justified by communitarian theory because neither fostersa community that is distinct from society at large and thatstill respects the bounds of the individual.

This is what communitarian theory requires. Communi-tarian theory envisions a political universe composed of allthree elements-self, community, and society-with commu-nity mediating between the other two. 274 If society and com-munity are identical, then the concept of community losesmeaning and offers the individual no refuge or mediatingcontext from which to relate to the larger society.275 Thedualist universe is reinstated with all its dangers. On theother hand, if community and individual are not differenti-ated, the individual is submerged, and the concept of com-munity, as Sandel and other communitarians conceive it, isundermined. The community that Sandel describes de-pends on the "reflective self-understanding" of its partici-pants for its formation and existence, and reflective self-understanding requires a core self to do the reflecting.276

271. See supra note 241 and accompanying text. For a definition of social insur-ance, see supra note 106, which also distinguishes social insurance from other in-surance mechanisms that involve group rather than societal responsibility.

272. See supra notes 54-56, 92-95 and accompanying text.273. See supra notes 52-65, 92-95 and accompanying text. Many would argue

that this already happens in many cases through the insurance process. Neverthe-less, this result is neither open nor obvious. See supra note 123.

274. See supra note 245; see also R. BELLAH, supra note 188, at 212, 251; Macneil,supra note 164, at 944-45. But see M. SANDEL, supra note 15, at 146.

275. See Macneil, supra note 164, at 936 n.141.276. See M. SANDEL, supra note 15, at 143, 150, 154-61. According to Sandel,

community is formed by acts of conscious self-awareness on the part of individu-als who realize and experience their attachments to others. This reflective con-sciousness is precisely why the formation of community expands and constitutes

1552 [Vol. 33:1473

BETWEEN TWO WORLDS

The paradox of the communitarian vision is that com-munity is necessary to constitute full persons, but individualpersons are necessary to build conscious, and therefore real,community. 277 Because it requires respect for the individualidentity of the person, as well as recognition of his associa-tional identity or community, communitarian theory doesnot represent a total break with the liberal theory, but rathera revision of it.278 Communitarian theory supports groupresponsibility, but it does not support societal responsibilityor the elimination of individual responsibility. 279 Can courtsand legal scholars actually structure such a form of responsi-bility in practice? Can any of the group causation theoriesthus far developed be so described? This brings us back tothe crucial problem of articulating an adequate definition forthe term "group" or "community."

Can courts hope to succeed where the political theoristshave failed? Courts that refuse to apply group responsibilityin any situation probably fear that they would be unable todraw lines and would gradually be led to eliminate individ-ual responsibility completely. Nonetheless, courts are ex-perts at drawing lines in concrete situations on a case-by-case basis. The common-law judicial process, not politicalphilosophy, may be the best hope for this difficult defini-tional task. Unfortunately, the courts that have adoptedgroup causation theories have only intuited the communitar-ian vision; therefore they simply have not addressed the defi-nitional question. In contrast, courts that have rejectedgroup causation have been paralyzed by the question.Neither result is desirable. Courts should recognize groupcausation; but they must also articulate its applications andlimits by specific reference to the meaning of group or com-

the self, which is one of the goods of community. Self-transcendence-which isone way to describe both the process and the good of community-requires afundamental or residual self. Otherwise, self-transcendence results in self-de-struction. When the self, as the source of consciousness and reflection, is de-stroyed, the continued life of the community through reflective self-understanding is also extinguished.

277. The communitarian vision differs from the social order of the premodernera. At that time, community identification was the product not of conscious self-awareness but rather of unconscious acceptance of circumstance.

278. See Macneil, supra note 164, at 936 n.136, 944-45.279. Adopting societal responsibility or eliminating individual responsibility

could be justified only by some form of welfarist, collectivist theory that woulddiffer from, and violate, both the communitarian and the liberal conceptions.

1986] 1553

UCLA LAW REVIEW

munity membership. As this Article has demonstrated, it isonly on the basis of its effect on community developmentthat group causation can be justifiably applied.

Some courts have already unconsciously suggested astarting point from which it is possible to build a definitionof group or community for the purpose of imposing groupresponsibility in tort law. For example, in Namm v. Frosst, aDES case involving multiple defendants, the court refused toimpose a form of group responsibility because it would re-sult in a taking of property "in order to pay for harm whichmay have been caused... by one.., who is unknown to thedefendants, over whom they have no control or even anymeaningful contact." 280 The court's language here suggeststhat it refused to apply group responsibility precisely be-cause it believed there was no meaningful group involved, asdefined by the parameters of mutual acquaintance, interac-tion, and influence. The justification for group responsibil-ity is encouraging group awareness and solidarity. If there isno significant potential for this result because the supposedgroup members lack any real or potential relationship, thenimposing group responsibility is senseless; it cannot accom-plish its purpose and can only weaken individual responsibil-ity unnecessarily.2 8'

Although the Namm court may not have used the abovereasoning, its language nevertheless suggests a startingpoint for a definition of group or community. When themembers of a putative injurer or victim group are situated sothat they either do have, or could potentially have, significantcontact and involvement with each other concerning thekind of activity or circumstance that is the subject of the dis-pute, then group responsibility would foster community for-mation and development and should be applied. At the veryleast, group responsibility should be considered and its com-munity-strengthening effects should be weighed carefullyagainst other competing concerns. 28 2

280. Namm v. Charles E. Frosst & Co., 178 NJ. Super. 19, 33-34, 427 A.2d1121, 1128 (App. Div. 1981).

281. SeeJ. FEINBERG, supra note 11, at 241.282. These competing concerns must include the desire to retain a cognizable

sphere of individual responsibility. Some consideration would have to be given tofactors indicating the extent to which such a sphere remains intact. This inquiry isbeyond the scope of this Article. In addition, to avoid de facto societal responsi-bility, the burden of responsibility should not be transferred so widely by the

1554 [Vol. 33:1473

1986] BETWEEN TWO WORLDS 1555

Applying such a "community viability" test would sup-port group causation in many of the cases that heretoforeadopted it on different grounds. In the DES cases, 283 a pro-ducer community is arguably viable, if not extant-despitethe contrary view of the Namm court, whose decision uncon-sciously suggested the community viability test but failed toapply it correctly. 284 In the asbestos cases,285 community vi-ability justifies imposing group responsibility on producers,but not employers. Employers are likely to be so differenti-ated and unconnected that no community is viable. In toxicwaste cases, 28 6 regionally defined waste producer communi-ties are potentially viable because all producers face similardisposal problems and could identify each other and inter-

group members that it would be distributed indiscriminately throughout societyas a whole. (This last point was raised by my colleague Professor LawrenceKessler.)

283. See, e.g., Tidler v. Eli Lilly & Co., 95 F.R.D. 332 (D.D.C. 1982); Morton v.Abbott Laboratories, 538 F. Supp. 593 (M.D. Fla. 1982); Mizell v. Eli Lilly & Co.,526 F. Supp. 589 (D.S.C. 1981); Ryan v. Eli Lilly & Co., 514 F. Supp. 1004 (D.S.C.1981); Sindell v. Abbott Laboratories, 26 Cal. 3d 588, 607 P.2d 924, 163 Cal.Rptr. 132, cert. denied, 449 U.S. 912 (1980); Payton v. Abbott Laboratories, 386Mass. 540, 437 N.E.2d 171 (1982); Abel v. Eli Lilly & Co., 418 Mich. 311, 343N.W.2d 164, cert. denied, 469 U.S. 833 (1984); Zafft v. Eli Lilly & Co., 676 S.W.2d241 (Mo. 1984); Namm v. Charles E. Frosst & Co., 178 N.J. Super. 19, 427 A.2d1121 (App. Div. 1981); Bichler v. Eli Lilly & Co., 79 A.D.2d 317, 436 N.Y.S.2d 625(1981), aft'd, 55 N.Y.2d 571, 436 N.E.2d 182, 450 N.Y.S.2d 776 (1982); Martin v.Abbott Laboratories, 102 Wash. 2d 581, 689 P.2d 368 (1984); Collins v. Eli Lilly& Co., 116 Wis. 2d 166, 342 N.W.2d 37, cert. denied, 469 U.S. 826 (1984); see alsosupra text accompanying notes 22-37.

284. See Namm, 178 N.J. Super. at 33-34, 427 A.2d at 1128. See supra notes280-81 and accompanying text.

285. See, e.g., Thompson v. Johns-Manville Sales Corp., 714 F.2d 581 (5th Cir.1983), cert. denied, 465 U.S. 1102 (1984); Hannon v. Waterman S.S. Corp., 567 F.Supp. 90 (E.D. La. 1983); In re Related Asbestos Cases, 543 F. Supp. 1152 (N.D.Cal. 1982); Starling v. Seaboard Coast Line R.R., 533 F. Supp. 183 (S.D. Ga.1982); Prelick v. Johns-Manville Sales Corp., 531 F. Supp. 96 (W.D. Pa. 1982);Hardy v. Johns-Manville Sales Corp., 509 F. Supp. 1353 (E.D. Tex. 1981), rev'd,681 F.2d 334 (5th Cir. 1982); Copeland v. Celotex Corp., 447 So. 2d 908 (Fla.Dist. Ct. App. 1984), rev'd in part and modified, 471 So. 2d 533 (Fla. 1985); see alsosupra note 149.

286. See Note, "Close Encounters of the Toxic Kind"--Toward an Amelioration of theSubstantive and Procedural Barriers for Latent Toxic Injury Plaintiffs, 54 TEMP. L.Q. 822,824-25, 831-36 (1981) (describing a number of documented case histories oftoxic waste spills and leaks). Few toxic waste cases have reached the appellatecourts because the doctrinal barriers to recovery under traditional torts rules, in-cluding the traditional causation requirement, are practically insurmountable. SeeGinsberg & Weiss, supra note 89, at 874-929; Note, supra, at 824-43. A groupcausation rule would greatly increase toxic waste victims' chances for success inrecovering damages as a group from waste producers.

UCLA LA W REVIEW

act. Note that potential community viability controls in thistest: a group of injurers could not escape liability by con-sciously avoiding interaction with each other. As for victimgroups, many populations exposed to toxic substanceswould seem to qualify, 287 because the circumstances that ledto their exposure-residing in the same area or working insimilar occupations and locations-support communityviability.

The community viability test would also define the lim-its to the application of group responsibility. Consider atoxic exposure case in which victims were exposed to carcin-ogens by several producers involved in completely differentactivities that used altogether different materials andprocesses. Robinson, who hypothesizes such a case, 288 ar-gues in favor of proportional recovery in these circum-I

stances. Victim community viability may exist in thissituation, if the exposures were concentrated in time andspace. However, no producer community is viable, becauseby hypothesis the producers' activities are so unrelated thatthere is no basis for interaction and mutual influence. Con-sequently, imposing group responsibility may not be justi-fied by communitarian reasoning. The case for groupresponsibility is even weaker if the exposures were diffusedover time and space, so that there is no viable victim com-munity either. The "cigarette cancer" 289 cases may present

287. See Jackson v. Johns-Manville Sales Corp., 727 F.2d 506 (5th Cir. 1984),reh'g en banc, 750 F.2d 1314 (1985) (reinstating portions of the panel's opinion,superseding portions of the panel's opinion, and certifying certain factual issuesto the Mississippi Supreme Court); Allen v. United States, 588 F. Supp. 247 (D.Utah 1984); Arnett v. Dow Chem. Co., 6 Chem. & Radiation Waste Litig. Rep.(BNA) 383, 390 (Cal. Super. Ct. 1983); Garner v. Hecla Mining Co., 19 Utah 2d367, 431 P.2d 794 (1967); see also supra text accompanying notes 38-65; supranotes 124-29 and accompanying text.

288. See Robinson, supra note 12, at 750-55.289. In the "first wave" of suits by smokers against cigarette manufacturers in

the 1950's, plaintiffs invariably lost, and statistical evidence of causation of cancerby smoking was excluded or found inadequate. See Note, supra note 127, at369-73 and cases cited therein. In recent years, a "second wave" of suits againstcigarette manufacturers has begun, the outcome of which is as yet uncertain,although the first case to go to a jury resulted in a verdict for the defendant. SeeGalbraith v. R.J. Reynolds Indus., No. 144,417 (Cal. Super. Ct., Sta. Barb. County1985). In this case, statistical evidence of causation was also excluded, and the

jury ruled for the defendant because, according -to the foreman, the "evidencewasn't there." Chambers, Tobacco Companies Breathe a Bit Easier, N.Y. Times, Dec.29, 1985, at E16, col. 4; see also Roysdon v. R.J. Reynolds Indus., 623 F. Supp.1189 (E.D. Tenn. 1985); Cippollone v. Liggett Group Inc., 593 F. Supp. 1146

1556 [Vol. 33:1473

BETWEEN TWO WORLDS

exactly this sort of situation. In these cases, victims arewidely scattered and possible injurers include not only to-bacco companies but also unrelated producers of environ-mental carcinogens.

Another important limit on group responsibility is thecase of traditional ambiguous causation, where a particularvictim's injury may have been caused by any of a number ofwholly unrelated actors, and some evidence (but less than apreponderance) points to every actor.290 Robinson's riskcontribution theory, like others that suggest using probabil-istic evidence and proportional recovery, would grant recov-ery here.29' Group causation-as justified and limited bycommunitarian theory-would reject recovery because noinjurer community is viable, and imposing liability on anyactor would violate individual rights solely, and unjustifiably,for collectivist reasons of deterrence and compensation.This result reemphasizes the point that the goal of imposinggroup responsibility, according to communitarian logic, isnot to increase deterrence or compensation, but to en-courage community formation as a distinct goal. When thisgoal cannot be furthered because community viability islacking, group responsibility is simply not justified undercommunitarian theory.

Perhaps the most difficult case to resolve is that whichinvolves a viable injurer or victim community as well as aspecific injurer and victim who are unequivocally identifi-able. For example, a labeled product, similar or identical toothers produced by different makers, injures a particular vic-tim who belongs to an identifiable user class that is exposedto, but whose members do not all suffer, a similar harm.The controversy over the "Dalkon Shield" contraceptive de-vice presents a case in point.292 Should group responsibility

(D.N.J. 1984). See generally Brody, Recovery Against Tobacco Companies, TRIAL, Nov.1985, at 48; Legal Times, Dec. 9, 1985, at 8, col. 2; Sitomer, Cigarettes on Trial:Tobacco Road is Strawn With Liability Lawsuits, Christian Sci. Monitor, Dec. 5, 1985,at 41, col. 1; Mintz, Cigarette Product-Safety Trial Begins, Wash. Post, Nov. 19, 1985,at CI, col. 4.

290. See supra notes 92-95 and accompanying text for a concrete illustration ofsuch a case.

291. See id.292. See, e.g., In re Northern Dist. of Cal. Dalkon Shield IUD Prods. Liab. Litig.,

521 F. Supp. 1188, modified, 526 F. Supp. 887 (N.D. Cal. 1981), vacated and re-manded, 693 F.2d 847 (9th Cir. 1982), cert. denied, 459 U.S. 1171 (1983); Rosenfeld

19861 1557

UCLA LA W REVIEW [Vol. 33:1473

be applied to the injurer or victim class here, even thoughthe known injurer will pay less than one hundred percent ofdamages to his known victims? The community formationand individual autonomy rationales clash most strongly inthis case. However, it should not be a foregone conclusionthat individual responsibility must prevail. The interests andbenefits of community formation may be significant enoughto override concern for the specific injurers' or victims' indi-vidual rights or responsibilities. 293

However, the purpose of this Article is not to specifyoutcomes for different cases. 294 The standard of community

v. A.H. Robins Co., 63 A.D.2d 11, 407 N.Y.S.2d 196, appeal dismissed, 46 N.Y.2d731, 385 N.E.2d 1301, 413 N.Y.S.2d 374 (1978).

293. On one level, imposing group responsibility in such a case could serve theliberal goal of individual fulfillment even better than the IR principle. For exam-ple, in a strong community, actual and potential victims could regain a sense ofdignity and connection by supporting each other. At the same time, they could begreatly empowered to confront and bargain with the corporate entities that affecttheir lives and well-being. On another level, the formation of a strong victim com-munity could encourage reflective and self-transcendent conduct, as well as theaffirmation and practice of unqualified virtues like loyalty, trust, charity, and thelike.

294. A number of other observations regarding specific decisions and doc-trines do follow from the argument in the text. First, if the purpose of legal rulesis to encourage community, then parties liable under group causation should notbe able to exonerate themselves by proving their actual causal innocence, a fre-quent result under the existing group causation decisions. See Sindell v. AbbottLaboratories, 26 Cal. 3d 588, 612, 607 P.2d 924, 937, 163 Cal. Rptr. 132, 145,cert. denied, 449 U.S. 912 (1980); Martin v. Abbott Laboratories, 102 Wash. 2d 581,605-06, 689 P.2d 368, 382-83 (1984); Collins v. Eli Lilly & Co., 116 Wis. 2d 166,197-98, 342 N.W.2d 37, 52, cert. denied, 469 U.S. 826 (1984). Even a causallyinnocent producer should be held liable if he could have influenced others whoactually caused the harm, since this would reinforce the incentive for community.The only ground for exoneration, then, would be lack of real ties to the responsi-ble group, or lack of responsibility on the part of the group itself.

Second, apart from the causation issue, a communitarian theory of law wouldsupport the view that the negligence issue should also be evaluated on a groupbasis. See supra note 96. Limiting the group responsibility rule to cases in whichall group members are negligent, as in the DES cases, is inappropriate. See, e.g.,Sindell, 26 Cal. 3d at 610-11,607 P.2d at 936, 163 Cal. Rptr. at 144. If any individ-ual within a cognizable community is negligent, then holding other members re-sponsible is justifiable because it would encourage stronger community bonds.The decision in Sheffield v. Eli Lilly & Co., 144 Cal. App. 3d 583, 192 Cal. Rptr.870 (1983), see supra notes 1, 119, was therefore misconceived, and was possibleonly because the Sindell court did not recognize or articulate the communitarianbasis for its original group responsibility approach.

Third, my comments do not necessarily imply that adoption of group respon-sibility would eliminate the negligence issue. Before imposing liability it couldstill justifiably be required that at least one group member be found negligent, orthat the riskiness of the group's activity as a whole be found excessive. On the

1558

BETWEEN TWO WORLDS

viability would evolve from the litigation process. In effect, Ipropose to create an issue of group cognizability similar tothat presented in class actions, although the determinativefactors would be quite different. In every case, either plain-tiff or defendant would have the prerogative to argue thatthe court should recognize an injurer or victim group andthen apply some form of group responsibility. This wouldlead to litigation about how such groups 'should be defined.It would force the courts to determine the appropriate defin-ing factors in light of the ultimate community solidarity ra-tionale. The legal process itself would help to answer thedefinitional question and to clarify the meaning of a "viablegroup" or "community." This suggestion implicitlypresumes that if a viable injurer or victim group, whetheractual or potential, could not be shown, liability could bebased only upon individual responsibility, whatever the con-sequences with respect to the burden of injurer liability orthe adequacy of victim compensation. The party seeking toinvoke group responsibility would have the burden of show-ing its appropriateness. Given the embryonic stage of ourunderstanding of the factors that define group viability andjustify group responsibility, this seems the most sensible ap-proach. Proposals like Robinson's that suggest a generaluse of probabilistic, proportional recovery would be rejectedbecause they make group responsibility the presumption.295

It is possible to look back now at the group causationdecisions and proposals presented in Part I and see that,while none show any explicit awareness or use of communi-tarian reasoning, some at least implicitly move toward it and

other hand, this requirement would assume that control of risk is the sole basis forgroup interaction and relation, whereas interaction over such issues as distribu-tion and handling of inevitable losses might also provide the basis for communitystrengthening. If so, then placing liability on a viable injurer community wouldmake sense even in the absence of any negligence. However, in the absence of aviable injurer community, both liability without fault and liability without identifi-able causation would be difficult to justify under communitarian theory. Thus thecommunitarian theory adds a new dimension to the debate over negligence andstrict liability, which has generally been conducted exclusively in corrective justiceand utilitarian terms. See, e.g., Calabresi & Hirschoff, supra note 96; Epstein, ATheory of Strict Liability, 2J. LEGAL STUD. 151 (1973); Posner, Strict Liability: A Com-ment, 2J. LEGAL STUD. 205 (1973).

295. See supra text accompanying notes 54-65.

1986] 1559

UCLA LA W REVIEW

reach results consistent with it. The Sindel1296 decision, byfocusing on the total fungibility of the product involved, ef-fectively narrows its application of group responsibility tocases in which producer community is viable.297 The AgentOrange298 opinion addresses the victim community issue inalmost explicit terms.2 99 The court decisions that adoptgroup causation are generally limited to situations involvingviable injurer or victim communities. On the other hand,the scholarly proposals 300 often call for results inconsistentwith the community viability test. These proposals seem tolack the courts' intuitive perception of the link betweengroup causation and the communitarian vision. Instead, theproposals tend to merge into a de facto social insurance ap-proach justifiable only on pure welfarist grounds.

Even though the courts deserve credit for their intuitionof the communitarian rationale for group causation, intui-tion is no longer adequate. By justifying their decisions interms of increased compensation and increased deterrence,and by failing to grasp the communitarian reasoning clearlyand make it explicit, the courts have confused group causa-tion with social welfarist compensation. In turn, this confu-sion has emerged in the commentary on these cases.Without clarification, the group causation decisions seem tosupport abandoning the IR principle for essentially collectiv-ist reasons. Unless courts go further and explicitly articulatethe communiitarian theory of group causation, they will faileven if they succeed. In the absence of the communitarianjustification and its self-imposed limits, group responsibilitycan degenerate into collectivism and submergence of the in-dividual. Even the most ardent supporters of the Sindell de-cision probably would not see these consequences as atriumph.

As this Article has demonstrated, both the shift to groupresponsiblity and the resistance it has encountered reflect,on one level, a fundamental tension between competing lib-

296. 26 Cal. 3d 588, 607 P.2d 924, 163 Cal. Rptr. 132, cert. denied, 449 U.S. 912(1980).

297. See id. at 610-11, 607 P.2d at 936, 163 Cal. Rptr. at 144; see also supra notes33-37 and accompanying text.

298. In re Agent Orange, 597 F. Supp. 740 (E.D.N.Y. 1984).299. See id. at 857-61; see also supra notes 83-88 and accompanying text.300. See supra text accompanying notes 38-65.

1560 [Vol. 33:1473

BETWEEN TWO WORLDS

eral and welfarist world views that extends-far beyond theconfines of tort law. On a different level, the shift to groupcausation can be seen as part of a larger movement towardan alternative, communitarian world view that in some mea-sure transcends the liberal-welfarist dichotomy. The courts'decisions thus reflect the larger political visions of societythat compete for our allegiance. At the same time, only byexplicitly recognizing and articulating the underlying frame-work of political visions and values can the courts make deci-sions that truly express and effectuate these visions, insteadof obscuring and distorting them. Nowhere is this truerthan in causation doctrine in torts. Clarity about the under-lying theory will produce sounder, and more soundly rea-soned, decisions. In addition, because the language andconceptual structure used by courts can reinforce or alterour political visions themselves, judicial clarity can play asignificant part in actually shaping the emerging vision ofcommunitarian political theory.

Therefore, I conclude that courts should apply groupcausation and responsibility theories and should explicitlyjustify them by the communitarian rationale. Moreover, thisapplication can and should depend explicitly on the issue ofgroup viability-the only factor directly related to the com-munitarian rationale. Courts should not, as they have donein the past, adopt group causation and responsibility theo-ries on ambiguous grounds. Nor should they rely upon thecollectivist grounds of deterrence and compensation, be-cause these are non-self-limiting and dangerous grounds onwhich to abandon the IR principle and its underlying valueof individual dignity. Relying on collectivist grounds canonly lead to unsound judicial decisions that lend support toan unsound collectivist political vision.

E. Postscript: A Communitarian Theory of Law

The communitarian theory has implications for legaldoctrine well beyond the issue of causation and the law oftorts in general. By articulating a vision of society in whichgoods, and goals, are defined in significantly different termsthan those of liberalism or social welfarism, communitarianthought adds a new dimension to jurisprudence. Tort theo-rists have argued for nearly two decades over whether legalrules make sense in terms of social welfarist criteria of

1986] 1561

UCLA LA W REVIEW

wealth maximization or liberal criteria of individual auton-omy.30 1 This controversy has carried over to other legalsubjects as well.30 2 By changing the terms of the discussion,a communitarian theory of law would open many new doorsthrough which to understand and criticize the function andmeaning of specific legal rules and doctrines. 303

In addition, at a more general level, communitarian the-ory casts new light on the debate over what is the best de-scriptive and normative theory of tort law and law ingeneral. On one level, communitarian theory seems to sup-port the emphasis of liberal, corrective justice theory on in-dividual rights and fulfillment, as opposed to the emphasisof welfarist, economic deterrence theory on overall societalwelfare. At the same time, however, communitarian theoryshows that in many cases where these theories were thoughtto be irreconcilable, they are not. Imposing group responsi-bility on a viable community can both further individual dig-nity, by strengthening a community essential to theindividual, and increase social welfare, by increasing deter-rence of injurious activity. Even though communitarian the-ory rejects welfarist goals such as deterrence andcompensation as ends in themselves, it accepts and even fa-cilitates them as by-products of the achievement of the com-munitarian goals of strengthening community andenhancing individual dignity. Communitarian theory revealsthat there is less inconsistency between liberal and welfaristgoals, and more potential for satisfying both, than is usuallyrecognized. 304

Thus, apart from suggesting a distinct and potentially

301. See, e.g., G. CALABRESI, supra note 3; Calabresi & Hirschoff, supra note 96;Epstein, supra note 294; Epstein, Causation and Corrective Justice: A Reply to Two Crit-ics, 8 J. LEGAL STUD. 477 (1979); Fletcher, supra note 3; Posner, supra note 294;Posner, The Concept of Corrective Justice in Recent Theories in Tort Law, 10 J. LEGALSTUD. 187 (1981); see also Englard, supra note 3; Wright, supra note 3.

302. See generally Posner, A Reply to Some Recent Criticism of the Efficiency Theory ofthe Common Law, 9 HOFSTRA L. REV. 775 (1981); Symposium on Efficiency as a LegalConcern, 8 HOFSTRA L. REV. 485 (1980); A Response to the Efficiency Symposium, 8 HoF-STRA L. REV. 811 (1980).

303. See supra note 294 for a discussion of how communitarian theory can beapplied to specific legal rules other than group causation.

304. This mediative aspect of the communitarian theory is entirely consistentwith the theory's concept of community itself as the crucial mediating structurebetween the individual and the society as a whole. See supra notes 225, 245 andaccompanying texts, supra text accompanying notes 274-75.

1562 [Vol. 33:1473

1986] BETWEEN TWO WORLDS 1563

fruitful new theory of law, the communitarian theory sug-gests a way to integrate existing theories and values that thusfar have been seen as irreconcilably opposed. It shows thatthe dichotomous world of liberalism and social welfarism,with its Hobson's choice between arid individualism and op-pressive collectivism, does not exhaust the field of possibil-ity. The communitarian world of self and society mediatedby community offers a new hope of integration. At present,we are still between two worlds. But stirrings all around us,in law and in other fields, suggest that we will not be sus-pended here much longer.